March-June 1999: An Overview
On March 24, 1999, the eyes of the world turned to Kosovo as aircraft from the North Atlantic Treaty Organization began to bomb targets in the Federal Republic of Yugoslavia. The start of NATO's air war against Yugoslavia was also the beginning of the bloodiest period in Kosovo since the end of the Second World War. In the twelve weeks that followed, Serbian and Yugoslav military, police, and paramilitaries expelled more than 850,000 ethnic Albanians from Kosovo, internally displacing several hundred thousand more.1 Many were robbed and beaten as they were forced from their homes, which were frequently looted and burned. Scores of women were raped. Thousands of adult males were detained, and many of them were executed, in some cases together with women, children, and the elderly, although the total number of civilians executed is still unclear (see section on Death Toll below). In more than a dozen mass killing sites, government forces tried to hide the evidence by destroying or removing bodies. The brutal campaign against ethnic Albanian civilians came to a halt only after the withdrawal of Yugoslav soldiers and Serbian police and paramilitaries and the entry of NATO forces on June 12, 1999.

Many observers mark the date of the NATO air war as the beginning of the Serbian and Yugoslav campaign. While March 24 saw a marked intensification of the campaign, the start of the operation actually came four days earlier, on March 20, when the monitors of the OSCE Kosovo Verification Mission (KVM) withdrew en masse from Kosovo. Most of the remaining international nongovernmental organizations evacuated their personnel at the same time. The departure of the KVM, together with international aid workers, deprived Kosovo not only of some of its most important witnesses, but also ended any deterrence that the presence of the OSCE verifiers might have provided.2 According to the OSCE report on its work in Kosovo from October 1998 to June 1999, based in part on interviews with refugees during the NATO bombing, "the level of incidents of summary and arbitrary killing escalated dramatically immediately after the OSCE-KVM withdrew on March 20."3

As the Background chapter of this book demonstrates, the abuses after March 20, 1999, were a continuation and intensification of the attacks on civilians, displacement, and destruction of civilian property carried out by Serbian and Yugoslav security forces during 1998 and the first months of 1999. By March 1999, the combination of fighting and the targeting of civilians had left an estimated 1,500-2,000 civilians and combatants dead.4 More than 200,000 Albanian civilians were internally displaced, almost 70,000 Albanians had fled the province to neighboring countries and Montenegro, and a further 100,000 Yugoslav nationals, mostly Kosovar Albanians, had sought asylum in Western Europe.5 Thousands of ethnic Albanian villages in Kosovo had been partially or completely destroyed by burning or shelling.6

The state-organized campaign that began in late March 1999 was nevertheless different in scale and scope than the violence that had occurred in 1998 and early 1999. Earlier operations by Serbian and Yugoslav security forces were directed at areas and families in Kosovo with ties to the KLA. Although the actions undertaken by Serbian police and Yugoslav soldiers in the Drenica region and southwestern Kosovo were egregious (in that they targeted civilians, illegal under international humanitarian law), they could be understood as brutal counterinsurgency against the KLA rebels.

The operations that began in late March 1999 went far beyond counterinsurgency: Serbian and Yugoslav forces carried out a systematic campaign of violence and forcible depopulation that left an estimated 80 percent of the civilians displaced from their homes.7 Areas with no history of support for the KLA and which had previously escaped the violence in Drenica and southwestern Kosovo, such as Pristina and eastern Kosovo, were targeted for mass expulsion. The killing and terror against civilians began to encompass any area with a current or historic link to the KLA, as well as some areas without any such link. In short, localized counterinsurgency was joined by systematic "ethnic cleansing."

Despite the scale of the displacement during 1998 and early 1999, many observers believed Kosovo, with its 90 percent ethnic Albanian population, would be exempt from large-scale ethnic cleansing, if only for the practical obstacles to the expulsion of an entire people. In ethnically-mixed Bosnia and Herzegovina, and contested areas of Croatia where no one ethnic group had an absolute majority, the expulsion of one ethnic group was a means of consolidating control over that territory by a rival group. By contrast, Kosovo with its overwhelming ethnic Albanian majority had experienced a steady outflow of its Serb population over preceding decades, with Belgrade resorting to forced resettlement of Croatian Serbs in an attempt to reverse the migration of Serbs out of the province. The slow initial response of UNHCR and NATO to the human tide of refugees into Macedonia and Albania in late March and early April is evidence that few in the international community believed the government of Slobodan Milosevic would attempt the "ethnic cleansing" of Kosovo.8 Many observers also believed that Milosevic would quickly capitulate once NATO airstrikes began.

If the conventional wisdom among Western observers held that "ethnic cleansing" was unlikely in Kosovo, what is the explanation for the systematic mass expulsion of the province's Albanian population by Serbian and Yugoslav forces between March and June 1999? Was it a coordinated plan or a spontaneous reaction to the NATO bombing? Although only Slobodan Milosevic and his top aides know the real explanation for the "ethnic cleansing" of Kosovo, several strong theories have emerged.

First, evidence suggests that the Milosevic government began preparing a large-scale anti-insurgency campaign to crush the KLA back in September 1998, long before anyone suspected NATO airstrikes, when the summer offensive was coming to a close (see Background). After routing the KLA from some of its strongholds-and influenced by growing international criticism and the onset of winter-the government halted the offensive and accepted the KVM monitors.

Between October and December 1998, a number of key individuals in the security apparatus were either removed or promoted, including the dismissal of Jovica Stanisic, head of Serbian state security, and Momcilo Perisic, chief of the army's general staff. A loyalist, Gen. Dragoljub Ojdanic, replaced Perisic, and Col. Gen. Nebojsa Pavkovic was promoted to commander of the Third Army, which had responsibility for southern Serbia and Kosovo. These changes, along with others in the Yugoslav Army and Serbian Ministry of Internal Affairs, suggest that a new strategy on Kosovo was being devised.

December 1998 and January 1999 saw a gradual increase of security forces in Kosovo, in violation of the KVM agreement, but without serious criticism from the international community. The killing of forty-five ethnic Albanians in Racak provoked an international outcry, but the Serbian and Yugoslav buildup inside Kosovo continued unabated. According to the OSCE, paramilitary groups were set up in Kosovo in February to organize transportation routes for looted goods.9 The buildup continued during the negotiations in Rambouillet, France, including the arming of local Serbs. According to the OSCE, which still had its monitors in Kosovo, February 6-23 saw "a significant build up of VJ forces throughout Kosovo leading to the arming of civilians and the training of reservists, the arrival of anti-aircraft weapons, the digging of tank pits and the preparation of demolition explosives along key routes in from the south and an increase in military air activity."10

The first attacks on KLA strongholds in the rural areas began immediately after the KVM departed on March 20, and these involved indiscriminate attacks on villages. According to the OSCE, emphasis was also placed on the strategically important route to the west of Vucitrn that links Kosovo to Serbia proper.

The character and intensity of the campaign appeared to change, however, after the commencement of NATO bombing. On March 24 began the full-scale cleansing of cities, such as Pec in Kosovo's southwest, and later Pristina, as well as the burning of Djakovica's old town. March 24-26 saw an aggressive operation to secure the southwest border with Albania, which involved large-scale displacement accompanied by killings of civilians.

While this evidence suggests that a powerful anti-insurgency campaign had long been in preparation before the NATO bombing, it seems that the Milosevic government took advantage of NATO air strikes to further impose terror on Kosovo's Albanian population and to expel large numbers of ethnic Albanians from the province. Using the pretext of the NATO bombing, the government was free to unleash a full-scale offensive on the KLA as well as to order the expulsion of more than 850,000 Kosovar Albanians.

One explanation is simply revenge. As hundreds of refugees testified, government forces repeatedly told them to "Go to NATO!" Unable to strike back at NATO air power, government forces took their vengeance on the civilian population. But the systematic nature of the expulsions contradicts this theory; clearly, there was a well-conceived plan to "ethnically cleanse" large portions of the Albanian population. Revenge was nothing more than an added factor to motivate the troops.

One credible theory is that Belgrade intended permanently to alter the demographic composition of Kosovo by expelling a large proportion of Kosovar Albanians-a strategy that had occasionally been proposed by Serbia's far-right, including former Serbian Deputy Prime Minister and head of the Serbian Radical Party Vojislav Seselj.11 This explanation is supported by the multiple accounts from refugees arriving in Albania of document confiscation and destruction and the removal of car license plates at the border. This so-called "identity cleansing" documented by Human Rights Watch and other organizations gathering testimony in Albania strongly suggests an attempt by Belgrade to strip Kosovo Albanians of their citizenship and to frustrate future efforts to return home.12

A second explanation for the "ethnic cleansing" is that it was designed to destabilize the neighboring countries of Albania and Macedonia. In March 1999, the young state of Macedonia, with two million inhabitants, at least 25 percent of whom are ethnic Albanian, was widely perceived as unstable and, at the same time, as a pivotal country for regional stability. The mass influx of refugees from Kosovo could easily have disrupted the fragile ethnic balance, if not destablized the entire country.13

Although Albania did not have these ethnic tensions to contend with, its stability had been repeatedly threatened in the 1990s by political and economic upheaval. The influx of more than 400,000 refugees could have pushed the impoverished country into turmoil.

With luck, none of these scenarios played out. There can be little doubt that the arrival of some 260,000 refugees from Kosovo placed an enormous strain on relations between ethnic Macedonians and ethnic Albanians in Macedonia. The government's go-slow policy in admitting refugees during the spring of 1999, the occasional police violence against the refugees, and the government's periodic refusal to admit additional refugees14 created enormous resentment and anger among Macedonia's Albanians. Equally, the common cause shown by Macedonian Albanians for their Albanian neighbors from Kosovo fueled fears about succession among ethnic Macedonians. The restraint shown by the leadership of the Albanian party in the ruling coalition, the Albanian Democratic Party, as well as the NATO presence, certainly helped prevent open conflict in Macedonia in the spring of 1999.

In Albania, the refugees were accepted with open arms. Despite difficult economic conditions, especially in northern Albania, refugees were successfully accommodated in refugee camps and private homes. The country even experienced an economic boom due to the influx of foreign humanitarian aid organizations, NATO, and the media.

Destabilizing Macedonia and Albania may also provide a third explanation for the "ethnic cleansing"-to weaken the resolve of the NATO alliance to continue the air war. It was clear from the outset that some members of the alliance, such as Italy, France, and Greece, were apprehensive about the airstrikes. Belgrade may have hoped that the flood of refugees would convince governments and the public that the bombing had made the situation worse, if not directly caused the refugee flow. (The Yugoslav government repeatedly claimed that the refugees were fleeing NATO bombs.) If this was Belgrade's intention, it badly failed: the "ethnic cleansing" and the need to reverse it became the strongest justifications for NATO's actions and helped to galvanize support among the alliance and the public to continue the bombing. Extensive media coverage of the expulsions, invoking images of Nazi deportations, helped solidify NATO support, notably during periods of intense criticism over civilian casualties by NATO and the bombing of the Chinese embassy in Belgrade.15 The refugee crisis handed NATO its greatest evidence that this was a war in defense of human rights.

Some commentators have suggested that the forced expulsion of Kosovo's Albanian population was an attempt by Belgrade to create conditions for the partition of Kosovo, presumably areas bordering Serbia in the northern and eastern parts of the province to be retained by Belgrade. The existence of such a plan, however, remains unclear since the expulsions were generally heaviest in those areas with high KLA activity rather than along any of the lines proposed as possible partition borders.

Finally, the "ethnic cleansing" apparently served some militarily strategic objectives. Villages in areas where the KLA was active, such as around Podujevo, Malisevo, and Djakovica, were cleared because of the logistical support they were providing to the KLA. Since it was difficult to discern between combatants and civilians, government forces drove the entire populations from many of these areas, essentially "draining the sea," to better engage the rebels. The one exception is the Drenica region, where many civilians were prevented from leaving the area and used as human shields.

Expelling refugees to Macedonia and Albania might also have been intended to tie down NATO forces in those countries and to hamper a possible ground invasion. Likewise, the crisis in Albania might have been an attempt to disrupt supply lines and hinder incursions by the KLA, who were based in the northern part of the country.

The Geography of Abuses

Beyond the systematic "ethnic cleansing" of Kosovo, some parts of the province were disproportionately affected-suffering the mass execution of civilians, rape, torture, and the destruction of civilian property through arson and looting. Many of these areas had already witnessed violence during 1998 and early 1999. The documentation and analysis of the abuses in these towns and villages form the main body of this report. (See Statistical Analysis of Violations for a study of Kosovo's most impacted municipalities.)

Areas with historic ties to the KLA were hardest hit. The municipalities of Glogovac and Srbica in the Drenica region, the cradle of the KLA, were the scene of multiple massacres of civilians, as well as arbitrary detentions, use of human shields, and the destruction of civilian property (see Drenica Region). With the exception of a mass expulsion from Glogovac town, most Drenica residents were prevented from fleeing, or unwilling to leave the area. Mass killings, expulsions, and the destruction of civilian property were also common in the municipalities of Djakovica, Orahovac, Prizren, and Suva Reka in the southern border area, where many of the villages had historically supported the KLA. (There are no reports of mass expulsion or killings among the almost exclusively Muslim Slav population of the southern-most municipality of Gora (Dragash).) With close ties to neighboring northern Albania, the southwestern region was the principal conduit for KLA weapons, supplies, and trained recruits, and it was across the Albanian border that hundreds of thousands of civilians were expelled over the course of a few weeks during the spring of 1999. The slaughter and terrorizing of civilians and the destruction of their homes in these municipalities between March and June 1999 can be understood, at least in part, as a continuation of the effort to eliminate the KLA's base of support in the civilian population that began in early 1998.

Explanations for the concentration of abuses in other municipalities investigated in detail by Human Rights Watch are more complex and less conclusive. The municipalities of Pec and Lipljan, both of which had significant Serb populations, were targeted for mass expulsion, but killings were more localized. Although the KLA was active in the Pec municipality and present in the western-most part of Lipljan municipality during 1998, there is little or no evidence to tie the KLA to some of the villages in which massacres occurred. The killings may simply have been designed to terrorize the remaining population into fleeing Kosovo, with the majority from Pec crossing into Montenegro and most of those from Lipljan being sent to Macedonia. Whatever the case, the frequent acquaintance between perpetrators and victims in these ethnically-mixed municipalities sets them apart from Drenica and southwestern Kosovo, where Albanians had scant familiarity with their tormenters. There was little KLA presence or violence during 1998 in the ethnically-mixed western municipality of Istok, for example. Nevertheless the municipality suffered mass expulsions of its Albanian residents into Montenegro and the burning and looting of their homes. Istok was also the scene of one of the bloodiest incidents of a bloody spring, when more than ninety Albanian inmates in the Dubrava prison were killed by unknown perpetrators in May 1999 after two days of NATO airstrikes that had already killed an estimated nineteen inmates (see Istok Municipality). By contrast, a massacre of more than one hundred civilians in the municipality of Vucitrn, as the villages around Sudimlja (Studime) were being forcibly expelled from Kosovo into Albania, appears almost _certainly tied to the presence of KLA fighters in the area. Many of Vucitrin's residents were also expelled, some toward Albania and others toward Macedonia.16

In other municipalities in Kosovo during the spring of 1999 (not covered in detail by this report), the picture is generally one of mass expulsion combined with more limited numbers of killings and other abuses. In northern Kosovo, the municipalities of Leposavic (Leposaviq), Zvecan, and Zubin Potok were relatively untouched, as reflected in the chapter on statistics, a fact mostly attributable to the majority Serb population in each.17 The Albanian-majority municipality of Kosovska Mitrovica (Mitrovice) was less fortunate. The ethnically-mixed and eponymous capital of the municipality, which has since become the most potent symbol of post-war ethnic division in Kosovo, was the scene of conflict between the KLA and Serbian security forces beginning in 1998.18 As elsewhere, the departure of the OSCE was followed by an increase in killings of Albanians and the forced expulsion to Albania of much of the town's Albanian population.19

Fighting between the KLA and Serbian security forces also occurred in the municipality of Podujevo (Podujeve) to the east of Mitrovica, although this region is less covered in this report. The violence and the accompanying crackdowns that began in September 1998 led some of its residents to flee to Pristina in December of that year.20 Tensions rose further in March 1999, notably in Podujevo town, with much of the remaining population expelled towards Macedonia or fleeing into the nearby hills, where a large number of displaced persons had taken shelter.21 The ethnically-mixed municipalities of Obilic (Obiliq) and Kosovo Polje (Fushe Kosove) had also witnessed some sporadic fighting during 1998, much of it linked to the seizure of a mine in the village of Grabovac (Graboc) in Kosovo Polje municipality.22

Following the departure of the KVM, events in both Kosovo Polje23 and Obilic24 followed a similar pattern, with much of the Albanian population forcibly expelled to Macedonia. The railway station in Kosovo Polje was also used as a transit point for the expulsion of tens of thousands of Albanians to Macedonia by train. The village of Ade (Hade) appears to have been singled out for punishment, perhaps because of an association with the KLA: five men from a single family in the village were executed and four of the corpses set on fire.25 A week later the entire population was forced to board buses and was expelled to Macedonia and Albania.26

The eastern municipalities of Klina (Kline) and Decani also suffered the forced expulsion of much of their Albanian populations during late March and early April. Ethnically-mixed Klina to the immediate west of the Drenica region had witnessed fighting between the KLA and Yugoslav and Serbian security forces during 1998. Much of the municipalities Albanian majority was expelled to Albania on March 31 and April 1, 1999.27 The predominantly Albanian population of Decani had also experienced KLA activity as well as substantial internal displacement prior to March 1999. Many of its inhabitants were expelled to Albania over the month of April 1999.28

March 1999 saw a broadening and deepening of the conflict in Kosovo to encompass areas that had previously experienced little violence and which had few links to the KLA. Kosovo's capital Pristina is a case in point. Although Pristina was spared the large-scale killing that occurred to the west in Drenica during the spring of 1999, a large proportion of the city's population was expelled towards Macedonia in the first wave of "ethnic cleansing" in late March and early April 1999. Robbery and looting were also widely reported, although far fewer properties were burned than in Pec or Djakovica. The eastern municipalities of Gnjilane (Gjilan) and Kosovska Kamenica (Kamenice) suffered a similar fate: although they largely escaped the killing and burning, thousands of their Albanian residents were expelled from their homes during April and May, with many robbed and detained on their way to the Macedonian border. (Little is known about the experience of the tiny Serb-majority municipality of Novo Brdo, which lies to Gnjilane's west.)

Municipalities close to the Macedonian border were not known for their links to the KLA, although the border was used by the group for smuggling weapons and supplies and was a focus of Yugoslav counterinsurgency efforts. Once the conflict broadened, however, Yugoslav and Serbian security forces moved quickly to expel or otherwise displace the population away from the border, with frequent reports that villages were being shelled and burned. Preparation for a NATO ground invasion may be one explanation.

The municipality of Kacanik, on the Macedonia border, had been mostly quiet during 1998. According to the OSCE KVM, the KLA moved into the area in February 1999, and armed clashes were reported between Yugoslav and Serbian security forces and the KLA in a number of villages during February and March.29 Efforts by the Yugoslav army and Serbian police to expel villagers to Macedonia began almost immediately following the departure of the OSCE, but many residents were reluctant to leave, leading to substantial and rapidly shifting internal displacement in the municipality before residents were forced into Macedonia in April 1999.30 There were also a number of killings in the municipality, notably in the village of Kotlina (Kotlino) where twenty-five men were executed (some of them members of the KLA) before the residents were expelled.31

The municipality of Urosevac (Ferizaj) to the north also experienced mass expulsion within weeks of the KVM departure. By mid-April, thousands of Urosevac residents were sheltering in camps in Macedonia, although a few were also forced to go to Albania.32 The railway line in Urosevac town made the municipality a convenient staging point for deportations from southern Kosovo into Macedonia. The ethnically-mixed border municipality of Strpce (Shterpce) experienced similar patterns of displacement to Kacanik, at least in its eastern half and other areas with Albanian populations. While initially resisting expulsion by moving from village to village, most of the municipality's Albanian inhabitants were in refugee camps in Macedonia by mid-April 1999.33 Many of the Albanian residents of Vitina (Viti) municipality, on the border to the east of Kacanik, were also expelled to Macedonia in early April 1999.

The Killings

Executions of ethnic Albanians took place throughout Kosovo. As the chapter on statistics makes clear, the vast majority of the victims were males, although females and children were not exempt. In numerous cases, such as the Vejsa household in Djakovica on April 1, the Gerxhaliu family in Donja Sudimlja (Studime e Poshtme) on May 31, or the Berisha household in Suva Reka on March 26, young children were killed along with adults. The statistics chapter also reveals how the killings generally occured in three distinct waves (see Graph 5).

Witness testimonies suggest three general motives for killings. The first was to expedite the "cleansing" process. Typically, security forces would shell and then enter a village, ordering the Albanian population to leave. A few individuals might be killed to spread panic and accelerate the deportation.

The second motive was to target individuals suspected of participating in or assisting the KLA. In village after village, security forces separated the men from the women and children, and interrogated the men about the insurgency, sometimes detaining them for days. As interviews with witnesses and survivors confirm, as well as statements in the international media given by Serbian and Yugoslav forces after the war, men suspected of KLA involvement were often shot on the spot. Such suspicion could be based solely on the fact that the man was of fighting age. Men were sometimes forced to strip, such as in Bela Crkva, to look for military uniforms under their civilian clothes, injuries from combat, or even dirt on their hands to suggest fighting. Since it was often difficult to distinguish civilians from KLA, not to mention village guards, government forces often took no chances, and killed those they thought might be a threat.

In other cases, prominent political leaders, community activists, or wealthy individuals were specifically targeted. As the chapter Forces of the Conflict makes clear, paramilitary units sometimes had lists of those to be killed, provided by local officials or the police.

A third factor was revenge. In certain cases, such as Meja near Djakovica or the Tusus neighborhood of Prizren, large-scale killings took place after the KLA had killed Serbian or Yugoslav forces. In Drenica, the killings in Vrbovac and Stutica occurred in late April immediately after NATO bombed the Feronikel plant, where security forces had been based.

Lastly, while most of the expulsions and killings were carefully planned, there are also cases where forces went out of control, especially volunteers or paramilitaries. The general lawlessness in Kosovo during the NATO bombing allowed for criminals and thugs to extort, rob, and kill with impunity. With very few exceptions, the government did nothing to stop these people. On the contrary, paramilitaries were dispatched to Kosovo despite their reputation for brutality, and some criminals were even released from prison if they agreed to go to Kosovo to fight.

Death Toll, The Missing and body removal

More than two years after the end of the war, the total number of victims killed between March and June 1999 remains unclear. Although the explanations for the lack of clarity in the death toll are straightforward and common to many post-conflict situations, the total number of dead remains one of the most controversial aspects of the war. Ultimately, however, what matters is not whether the dead number 5,000 or 15,000, but that large numbers of civilians were targeted for execution by Serbian and Yugoslav security forces.

One reason for the number controversy is the exaggerated claims made by NATO and NATO governments during the war. Some U.S. officials, including Secretary of Defense William Cohen and State Department Special Envoy on War Crimes David Scheffer suggested that up to 100,000 Albanian men were missing and feared dead.34 Such figures contrast with the more measured U.S. government and NATO estimates from the same period of between 3,000 and 4,000, based on refugee accounts.35 After the war, head of the U. N. administration in Kosovo, Bernard Kouchner, said that "around 11,000 people" had died, although his press office later backpedaled from that estimation.36 Still, the unproven claims by top government officials at the height of the war led to charges of propaganda to justify NATO intervention both by journalists and by NATO's political opponents in the West.37 These allegations also fueled predominantly left-wing critics of Western policy towards the former Yugoslavia, some of whom dispute that the mass exodus of Albanians was caused by Serbian government forces and even that mass killings of Albanian civilians occurred.

The more direct reason for the uncertainty, however, is a deliberate attempt on the part of the Serbian and Yugoslav government to destroy evidence and remove bodies. Both the ICTY and Human Rights Watch have documented cases where bodies were disinterred and removed from the crime scene, in an apparent attempt to conceal the killing. A radio documentary broadcast on National Public Radio in the United States on January 25, 2001, called Burning the Evidence, claims that Serbian and Yugoslav forces systematically transported the bodies of Kosovar Albanians to the mining complex at Trepca near Kosovska Mitrovica, where they were incinerated. Citing Serbian fighters and "a well-placed Serbian intelligence officer," between 1,200 and 1,500 bodies were destroyed at Trepca, according to the report.38

The credible allegations of body removal and destruction were apparently confirmed in May 2001, when the Serbian government announced that a truck filled with eighty-six bodies had been dumped in the Danube River in Serbia during the Kosovo war-allegedly the bodies of ethnic Albanian civilians taken from Kosovo. A top Serbian Ministry of Internal Affairs official announced on May 25 that Milosevic had ordered then-Interior Minister Vlajko Stojiljkovic to eliminate "all traces which could lead to any evidence of crimes" in Kosovo.39 As of July 2001, the Serbian authorities had announced the discovery of four additional graves in Serbia with as many as 1,000 Kosovar Albanian bodies.

As of July 2001, the ICTY had exhumed approximately 4,300 bodies, far less than the 11,334 bodies initially reported to the ICTY by Kosovar Albanians.40 According to the International Committee of the Red Cross, as of April 2001, 3,525 people remain missing from the conflict - the vast majority of them Kosovar Albanians.41

Between June and November 1999, ICTY teams exhumed 2,108 bodies from 195 grave sites. During the second phase of exhumations, between April and November 2000, forensic experts examined another 325 sites and found an additional 1,577 bodies and 258 incomplete remains. In her November 2000 address to the U. N. Security Council, ICTY Chief Prosecutor Carla Del Ponte stated that the provisional total of exhumed bodies was "almost 4,000 bodies or parts of bodies." It should be noted, however, that the ICTY apparently made no distinction between combatants and non-combatants, and that in some areas the KLA was present among the civilian population (eg: in the Vrbovac-Stutica area of Drenica during the April 30 offensive that left dozens of ethnic Albanians dead).

Of crucial importance, however, are the statements about grave tampering made by Del Ponte in both her 1999 and 2000 address to the Security Council. In 1999, after exhuming 2,108 bodies, she said:

This figure [2,108 bodies] does not necessarily reflect the total number of actual victims, because we discovered evidence of tampering with graves. There are a significant number of sites where the precise number of bodies cannot be counted. In these places, steps were taken to hide the evidence. Many bodies have been burned, but at those sites the forensic evidence is nevertheless consistent with the accounts given by witnesses of the crimes.42

In the next year's address, Del Ponte stressed that "it will never be possible to provide an accurate figure for the number of people killed, because of deliberate attempts to burn the bodies or to conceal them in other ways."43

Sites at which the ICTY has reported clear evidence of grave tampering include Izbica, Trepca, Mala Krusha and Crkolez (Cerkolez). Human Rights Watch, the OSCE, and other organizations documenting violations of humanitarian law and human rights have also collected evidence of grave tampering and other efforts to conceal evidence of killings by Yugoslav and Serbian forces prior to June 12, 1999. These include the removal of bodies, the reinterring of bodies from mass graves into individual graves, the burning of corpses, and the removal or exchange of clothing and personal effects in order to complicate the process of identification.

Human Rights Watch gathered credible testimony relating to grave tampering in Mala and Velika Krusa, Pusto Selo, Slovinje, Poklek, Kotlina, Rezala (Rezalle), Izbica, Trnje (Trrnje), and Djakovica, where at least seventy-seven bodies were removed from the town cemetery. In Trnje near Suva Reka, between twenty-four and thirty-six village men were killed by Serbian forces on March 25. Four days later, a witness saw unidentified men taking the bodies away on a truck. He said:

Early in the morning I heard a truck come from Leshane [Lesane in Serbian]. I heard them stop, and they opened the metal doors, and I knew they came to take the bodies. I heard when they put them in the truck, and I heard the Serbs complain about the smell. They put them all in, and went back to Leshane.44

In Slovinje, Human Rights Watch visited a temporary grave site outside the village in which the sixteen victims of an April 16 killing had been buried by their relatives the following day (see Lipljan Municipality). The bodies were later removed by Serbian security forces with excavation equipment, and the relatives have no information about the remains. In Poklek, the estimated forty-seven victims of an April 17 killing were first machine-gunned before hand-grenades were thrown into the room where they lay (see Drenica Region). Several other corpses were pulled from the garden well. Finally the house was set on fire. Relatives showed a Human Rights Watch researcher a box of human bones reportedly collected from the room during a July 1999 visit to the site. In the village of Kotlina, Human Rights Watch saw a wooded site where Serbian security forces had attempted to conceal evidence of the killing of twenty-two men by putting their bodies into two deep natural wells and dropping explosives into the holes.

There have been two scientific studies to address the question of how many Albanians were killed. The first study, released in June 2000 by the U.S. Centers for Disease Control,45 is an epidemiological analysis of all deaths in Kosovo during 1998 and 1999. It concludes based on its household surveys and pre-war mortality rates that an estimated 12,000 persons died as a result of "war-related trauma." It is important to emphasize, however, that the study includes 1998 (when approximately 1,500 Albanians were killed as a result of the conflict) and does not undertake the difficult distinction between civilians and combatants.

The second study, Political Killings in Kosova/Kosovo, published in October 2000 by the Central and East European Law Initiative (CEELI) of the American Bar Association and the American Association for the Advancement of Science (AAAS), concluded that approximately 10,500 Kosovar Albanians were killed between March 20 and June 12, 1999, with a 95 percent confidence interval from 7,449 to 13,627.46 The report further analyzes the timing and location of the killings, showing that the killings correlated closely with the flow of refugees out of Kosovo (see Statistical Analysis of Violations).

Targeted Killings

In addition to the general killings that took place throughout Kosovo, some prominent Kosovar Albanians were specifically targeted for execution. A number of politicians, doctors, human rights activists, and other public figures were killed in this way, as well as individuals who had worked with the OSCE or rented their homes to the organization.

The first reported case was the murder of a well-known human rights lawyer, Bajram Kelmendi, and his two sons, Kushtrim and Kastriot, aged eighteen and thirty-one. Bajram was active with the Council for the Defense of Human Rights and Freedoms and had been providing information during 1998 to the war crimes tribunal. He was highly respected for his skills as a defense attorney in numerous politically-motivated trials over the previous decade.47 Most recently, he had defended the Albanian-language newspaper Koha Ditore, which was on trial for publishing a statement by the KLA's political representative, Hashim Thaci.

On the night of March 24, the beginning of the NATO bombing, five uniformed policemen forced their way into the Kelmendi house on Vellusha Street in Pristina.48 According to Kelmendi's wife, Nekibe Kelmendi, herself a prominent lawyer, Kastriot called the police to tell them that someone had broken into their home, but the police hung up the receiver. Nekibe tried to call and they hung up on her as well. The family was forced to lie down, and Bajram was hit in the stomach. The police searched the house but didn't steal anything. Around 1:30 a.m., Bajram, Kushtrim, and Kastriot were taken away in Kastriot's car. Kushtrim was told that he should "kiss his children for the last time."

Nekibe tried without success to get information about her husband in the police stations and hospitals of Pristina. She also contacted Natasa Kandic, executive director of the Humanitarian Law Center, one of Yugoslavia's most respected human rights groups, who publicized the abduction. On the morning of March 26, Nekibe received word that the bodies of Bajram and her two sons had been found at a gas station on the Pristina road leading to Kosovo Polje. All three had been shot.

The murder of Bajram Kelmendi and his sons terrorized Pristina and Kosovar Albanian professionals-anyone with a public profile as a politician, human rights activist, journalist, artist, or academic. Panic spread in Pristina out of fear that unaccountable forces would liquidate prominent members of Kosovar Albanian society. Upon receiving news of Kelmendi's death, many Albanians decided to leave Kosovo.

Other prominent Albanians were killed in different parts of Kosovo around the same time. On March 25, Serbian security forces broke into the home of a respected physician in Djakovica, Dr. Izet Hima, and shot him in front of his wife. A well-known lawyer, Urim Rexha, and an LDK leader, Mark Malota, were also killed in Djakovica that day (see Djakovica Municipality). Two high level LDK activists in Mitrovica were also killed in unclear circumstances: Agim Hajrizi and Latif Berisha.

On March 27, unknown security forces reportedly took into detention the LDK secretary in Kosovo Polje, along with his wife and another Albanian family which lived nearby. According to the OSCE's report on Kosovo, the males were beaten badly in detention. They were released on March 28 and eventually made their way to Macedonia. On April 4, the LDK secretary, who is not named in the OSCE report, died from his wounds in a hospital in Tetovo, Macedonia.49

Perhaps the most prominent killing was the murder of Fehmi Agani, a professor and leading member of the LDK. Agani was a politician respected by most people involved in Kosovo politics: his own party, the KLA, international negotiators, and even his Serbian interlocutors. For many years, he was viewed as the key player behind LDK politics, and a person in genuine search of a peaceful solution to the crisis. His intelligence and popularity are a possible motive for his murder.

The circumstances behind Agani's death remain unclear. On May 6, Agani attempted to leave Kosovo by train with his family. On the border with Macedonia, the train was sent back. Somewhere near Kosovo Polje, police stopped the train and ordered everyone off. Various accounts in the press have Agani being taken away either in a bus with other Albanian men or separately with police in a private car. His body was found the following day next to a dirt road near Lipljan. The Serbian government claimed that the KLA had killed Agani to prevent him from negotiating between Rugova and the Serbian government. A statement from the police issued on May 7, 1999, said:

It is assumed that the terrorists of the so-called KLA kept Agani isolated in order to prevent his engagement in negotiations between Ibrahim Rugova and the Government of Serbia. When Rugova left for Rome they had no further interest to keep him, so they killed him. This most recent terrorist act can be, without any doubt, interpreted as a confirmation of an already announced decision of the KLA to do the same thing to Mr. Ibrahim Rugova.50

Other well-known individuals, such as Latif Berisha, the LDK president in Kosovska Mitrovica,51 Din Mehmeti, a poet, and Teki Dervishi, a writer, were also reportedly killed.52 On March 24, the guard at the newspaper Koha Ditore was shot and killed by police who were raiding the offices.

Another targeted group were Albanians who worked with the OSCE's KVM mission. Once the KVM withdrew from Kosovo on March 19, many of their offices were looted and burned, and some of the Albanian staff was sought by police or paramilitaries for beatings or execution. The worst case of revenge took place in Suva Reka against the Berisha family, which had rented two houses to the OSCE. Serbian forces killed at least twenty-four members of the family on March 25 and 26, including eleven children aged sixteen or younger (see Suva Reka Municipality).

Rape and Sexual Assault

Rape and other forms of sexual violence were used in Kosovo in 1999 as weapons of war and instruments of systematic "ethnic cleansing."53 Rapes were not rare and isolated acts committed by individual Serbian or Yugoslav forces, but rather were used deliberately as an instrument to terrorize the civilian population, extort money from families, and push people to flee their homes. Rape also furthered the goal of forcing ethnic Albanians from Kosovo.

In total, Human Rights Watch found credible accounts of ninety-six cases of sexual assault by Yugoslav soldiers, Serbian police, or paramilitaries during the period of NATO bombing, and the actual number is probably much higher. In six of these cases, Human Rights Watch was able to interview the victims in depth. Human Rights Watch met two other women who acknowledged that they had been raped but refused to give testimony. Finally, Human Rights Watch documented six cases of women who were raped and subsequently killed.

The ninety-six cases also include rape reports deemed reliable by Human Rights Watch that were compiled by other nongovernmental organizations.54 To the extent possible, Human Rights Watch corroborated these accounts through interviews with dozens of nurses, doctors, eyewitnesses, and local human rights and women's rights activists. It is important to note that some of these cases may have been double-counted by local and international organizations. Despite this, Human Rights Watch believes that the actual number of women raped in Kosovo between March and June 1999 was much higher than ninety-six, since Kosovar Albanian victims of rape are generally reluctant to speak about their experiences. At the same time, it should be noted that Human Rights Watch was not able to confirm the allegations of rape camps in Kosovo that were presented during the war by the U.S. and British governments, as well as by NATO.

In general, rapes in Kosovo can be grouped into three categories: rapes in women's homes, rapes during flight, and rapes in detention. In the first category, security forces entered private homes and raped women in front of family members, in the yard or in an adjoining room. In the second category, internally displaced people wandering on foot and riding on tractors were repeatedly stopped, robbed, and threatened by the Yugoslav Army, Serbian police, or paramilitaries. If families could not produce cash, security forces sometimes threatened that their daughters would be taken away and raped; in some cases, even when families did provide money, their daughters were taken away. The third category of rapes took place in temporary detention centers, such as abandoned homes or barns.

As an example of the second category, one rape victim recounted to Human Rights Watch how she was dragged off a tractor by a Serb paramilitary near the border village of Zur (Zhur) and sexually assaulted in front of dozens of other refugees. The victim, a thirty-year-old mother traveling with her mother, mother-in-law, and two children, told Human Rights Watch:

Two uniformed Serbian men stopped us. A big guy with red hair called me from the tractor. The red-haired one came around the tractor and said, "You," pointing at me. When he told me to get off the tractor, I didn't. Then he yelled, "You! Get off!" My three-year-old son was asleep on my lap. He kept yelling, "Get off! Get off!" He pulled me off the tractor and ripped my clothes. His pants were already open and his penis was out. He tore off my bra. I started screaming and crying. The other Serb came close and pointed his automatic weapon at my chest. I was wearing dimije [baggy pants] so they'd think I was old. The red-haired one took my pants off, tearing the drawstring. He told me to sit down. He took the 10 DM that I had with me. He took off his pants and pulled me close to him. We were right next to the tractor, next to the driver's cabin. I had my period. When he took off my pants, he saw the pads with blood on them, so he didn't have sex with me. Instead he turned me around and grabbed my breasts, trying me on the other side [anal rape]. I contracted myself very tightly and he didn't succeed. He may have ejaculated. I don't know. It took three or four minutes, then he told me that I could get back on the tractor.55

Witnesses to this attack, which occurred on June 2, 1999, corroborated the account and provided additional, credible details of the incident. A tractor driver who passed that same point later in the day, as well as his other passengers on the tractor, corroborated the description of the two uniformed men. One eyewitness to the sexual assault, an eighteen-year-old man from Djinovce (Gjinoc) in the Suva Reka municipality, told Human Rights Watch:

He took her onto the asphalt road and raped her right there in front of everyone. Only one Serb raped her. The other Serb hit people with the butt of his automatic weapon and said, "Silence, silence!" We all averted our eyes. It took three or four minutes. He did it right next to the tractor.56

With few exceptions, the rapes documented by Human Rights Watch were gang rapes involving at least two perpetrators. In several cases, victims and witnesses identified the perpetrators as Serbian special police, in blue or blue-camouflage uniforms, or Yugoslav Army soldiers, in green military uniforms. The majority of rape cases, however, were evidently committed by Serbian paramilitaries, who wore various uniforms and often had bandanas, long knives, long hair, and beards. These paramilitary formations worked closely with official government forces, either the Serbian Ministry of Internal Affairs or the Yugoslav Army, throughout Kosovo (see Forces of the Conflict).

The Serbian and Yugoslav authorities knew that their paramilitaries had used rape and other forms of sexual violence in Bosnia and Herzegovina. Yet, the paramilitaries were deployed to or allowed to operate in Kosovo by the Serbian and Yugoslav authorities apparently without any precautions being taken to prevent their committing further such war crimes.

The participation of Serbian and Yugoslav forces in gang rapes renders it unlikely that senior officers were unaware of the assaults. Rapes occurred frequently in the presence, and with the acquiescence, of military officers. Several rape victims actually reported the crimes to Yugoslav military officers. Yet there is no evidence that the Yugoslav Army or the Serbian Ministry of Internal Affairs made any attempt to apprehend or punish those responsible for the attacks. Despite this seeming dereliction of duty, many leading police and military officers from the Kosovo campaign have been honored or promoted within the Serbian and Yugoslav forces since the end of the conflict.

There is also no evidence that the Yugoslav Army or Serbian Ministry of Internal Affairs took any measures to prevent rape and other forms of sexual violence, such as issuing orders or warning troops that they would be punished for these crimes, although there were some cases where soldiers or police tried to protect women from paramilitaries. Moreover, soldiers, police, and paramilitaries often raped their victims in front of many witnesses. In addition to actual rapes that took place in front of others, the process of pulling women out of refugee convoys often occurred in full view of other internally displaced persons.

Forced Expulsions

During 1998 and the first months of 1999, there was substantial displacement in Kosovo: more than 200,000 Kosovar Albanians were internally displaced, almost 70,000 fled the province to neighboring countries and Montenegro, and a further 100,000 Yugoslav nationals, mostly Kosovo Albanians, sought asylum in Western Europe.57 But as the discussion above makes clear, few Western observers believed that Belgrade would attempt to expel the entire Albanian population of Kosovo. None could have predicted the speed and scale of the expulsion: within three weeks of the start of NATO airstrikes, there were 525,787 refugees from Kosovo in neighboring countries.58 A month later, on May 12, the total had risen to 781,618.59 All told, the Yugoslav military and Serbian police and paramilitaries expelled 862,979 Albanians from Kosovo,60 and several hundred thousand more were internally displaced, in addition to those displaced prior to March.61 These figures indicate that by early June 1999, more than 80 percent of the entire population of Kosovo and 90 percent of Kosovar Albanians were displaced from their homes. Approximately 440,000 refugees crossed the border to Albania and 320,000 to Macedonia (of whom almost 80,000 were transferred to third countries outside the region). Montenegro hosted around 70,000 refugees, while Bosnia and Herzegovina received more than 30,0000.

Refugee flows from Kosovo between March and June generally followed the principle of proximity to borders. Residents from the western half of Kosovo generally crossed the southwestern border into Albania and residents from the eastern half generally crossed the southeastern border into Macedonia. Conversely, residents of Kosovska Mitrovica, Vucitrn, and other ethnic Albanian areas in northern Kosovo were generally sent to the Albanian border. Many residents from Pec and the western-most part of Kosovo crossed into Montenegro, while some residents in the eastern-most municipalities entered first into southern Serbia before crossing the border into Macedonia. Some areas saw relatively few departures: A military cordon and an unwillingness to leave meant that few residents from the Drenica region in central Kosovo fled the province, other than those expelled from Glogovac town in early May (see Drenica Region). While the fact of proximity in most cases may suggest voluntariness, the statements of hundreds of refugees interviewed by Human Rights Watch and other organizations in Albania, Macedonia, and Bosnia indicate otherwise.

The flight of the ethnic Albanian population from Kosovo was executed with a degree of coordination and control that render it impossible to reach any conclusion other than systematic forced expulsion. At least four factors are key in determining that a policy of "ethnic cleansing" was carried out in Kosovo: First is the timing of the refugees' arrival-refugees arrived in Macedonia and Albania from the same areas on the same dates, and at various times (notably during negotiations) the flow of refugees stopped or was switched from one border to another. Second is the means of departure: refugees were expelled into Macedonia by train, which allowed the efficient removal of thousands of persons a day. Others, including many of those sent to Albania who did not have their own transportation, were taken by trucks and buses organized by the Serbian police. Collection points were used to facilitate expulsion. Third is the use of threats and violence to terrorize the population into departing, a central element of "ethnic cleansing," observed frequently during the wars in Bosnia and Croatia. Fourth is the practice of "identity cleansing": refugees expelled toward Albania were frequently stripped of their identity documents and forced to remove the license plates from their cars and tractors before being permitted to cross the border.

Since almost 90 percent of the refugees who left Kosovo crossed the Albanian or Macedonian border, the analysis of the timing and pattern of expulsion must necessarily focus on arrivals in those two countries. Human Rights Watch had researchers present in both countries from the last week of March until the end of the war in mid-June 1999. Researchers were therefore able to observe both the timing and pattern of refugee arrivals. If refugees were fleeing the NATO bombing or the fighting between Yugoslav and Serbian forces and the KLA, one would expect that refugees would arrive in a somewhat random fashion, based on an individual family's decisions about when the risks became too great to remain. In fact, almost all refugees arrived village-by-village and municipality-by-municipality in Albania or in Macedonia. The exceptions were cases where some residents from the same place were directed to Albania and some to Macedonia.

Although they were necessarily selective, the chronologies of Human Rights Watch interviews provide some insight into that process. In Macedonia, Human Rights Watch interviewed mostly newly arrived refugees from Pristina in the first week (beginning March 24). In the second and third weeks most of the new arrivals were from Kacanik and Strpce municipalities. In the fourth week, large numbers of refugees arrived from Gnjilane. In week five, many of the new arrivals were from Lipljan municipality. Similarly in northern Albania, Human Rights Watch interviewed large numbers of new arrivals from Prizren municipality during the first week; in the second week, many of the refugees came from Suva Reka and Djakovica, and in the third week, there was a large influx of refugees from Mitrovica and more from Djakovica. Refugee flows to Macedonia were abruptly stopped on several occasions, notably in early June, when the details of a settlement between the Yugoslav government and NATO were being negotiated. And as the OSCE noted, "the flow of refugees was also regulated, with the result that many thousands would arrive at the border crossing points with Albania, the former Yugoslav Republic of Macedonia and Montenegro (FRY) on some days, but then only a handful of refugees would arrive at particular crossing points on succeeding days."62

A more systematic effort to analyze the pattern of refugee flows into Albania, conducted by the American Association for the Advancement of Science (together with the East-West Management Institute and the Institute for Legal and Policy Studies in Albania), confirms this analysis. The statistical study primarily utilized data collected during the crisis by Albanian guards at the Morina border crossing in northern Albania. The study concludes that "the mass exodus of refugees from Kosovo [into Albania] occurred in patterns so regular that they must have been coordinated." The report further compared the expulsion times and patterns to the NATO bombing and found that "only a small fraction of Kosovo Albanians fled Kosovo as a direct result of NATO bombing raids."63 (See Statistical Analysis of Violations for a further discussion.)

The pattern of departure of many of the refugees also strongly points to organized expulsion rather than spontaneous departure. The depopulation of Kosovo's capital, Pristina, is a case in point. Within days of the start of the NATO airstrikes, Serbian police and paramilitaries began systematically to clear the city of large sections of its Albanian population. Witnesses from the neighborhoods of Vranjevac (Kodra e Trimave), Tashlixhe, Dragodan, and Dardanija (Dardania) told Human Rights Watch that police and masked paramilitaries went door to door at the end of March, telling residents that they had to leave at once. M.B., a mother of two from Tashlixhe said that she had been told: "Come on, get out! You must go to the railway station."64 In some cases witnesses were told they would be killed if they failed to comply. A medical doctor and his family were told by masked men "if you don't leave in one minute we will kill you all!"65

Upon leaving their homes, residents were directed by police towards the central railway station in Pristina, while others left by car. The side roads were blocked by armed police and paramilitaries: a Vranjevac resident said that "people who tried to walk in another direction were forced back by police."66 Thousands of Pristina residents were gathered at the railway station, with armed police posted around the area, where they were herded onto a passenger train headed for the Macedonian border. The trains were extremely overcrowded: one refugee said he was one of twenty-eight people forced into a compartment meant for eight passengers.67 Several refugees also described people being loaded onto buses and trucks at the railway station, which suggests that it served as a general collection point for the organized expulsion of ethnic Albanians from Pristina. Refugees interviewed by Human Rights Watch saw police drag people from cars and beat them. Most indicated that police and especially paramilitaries demanded hard currency for safe passage, and reported witnessing the theft of jewelry and vehicles. The use of collection points was also reported by refugees in the towns of Glogovac and at the central railway stations in Urosevac and Kosovo Polje, and in the village of Belanica (Bellanice) in Suva Reka municipality.

As the case of Pristina suggests, violence and threats were frequently used to terrorize the population into departing. Residents from Lipljan municipality told Human Rights Watch that the first massacre in the municipality (in the village of Slovinje) made them nervous. A climate of fear was created by subsequent killings in Malo Ribar (Ribar i Vogel) and Mali Alas (Hallac i Vogel), which left them with no choice but to leave (see Lipljan Municipality). Random gunshots, police checks, and house burnings also contributed to their decision to flee.

In areas with a history of support for the KLA, killings and other violence served the dual purpose of terrorizing the population and eliminating persons associated with the rebels. The experience of the town of Djakovica is indicative: a wave of organized terror early in the war left dozens dead and thousands as refugees in northern Albania (see Djakovica Municipality). This first wave of violence in Djakovica began on March 24, when NATO initiated its attack on Yugoslavia, and continued until April 2.

Following the principles of "ethnic cleansing" in Bosnia and Herzegovina, prominent residents, including doctors, lawyers, and politicians, were targeted for death-a strategy designed to terrorize the remainder of the population into believing that no one was safe and to eliminate important sources of leadership in the community. Approximately 200 Djakovica residents are believed to have been killed between March 24 and April 2, 1999, many of them in a series of house-to-house operations by Yugoslav soldiers, Serbian police, and paramilitaries.

Much of the population was expelled during this period. Witnesses interviewed by Human Rights Watch in northern Albania-most of them women- described their forced expulsion at the hands of police, paramilitaries, and the army. Many reported seeing clusters of dead bodies as they left the city.68 According to one Djakovica resident:

It was all very organized. They went from one neighborhood to another. Some soldiers are in charge of destroying things, while others are in charge of accompanying people to the border.69

The widespread confiscation of identity documents and car license plates by Serbian police and border guards from departing Kosovar Albanian refugees also points to the systematic nature of the expulsions. Hundreds of refugees arriving in Albania spoke of being forced to hand over ID cards, passports, and birth certificates, which were often torn up in front of them, before they were permitted to cross the border. Those who crossed the border by car were given screwdrivers and ordered to remove the license plates from their vehicles. By contrast, refugees who were expelled to Macedonia generally were permitted to retain their documents, even after having them inspected by Serbian police officers. (As noted in the section discussing explanations for the "ethnic cleansing," the difference in approach may reflect an expectation that those sent to Albania could be more easily characterized as Albanians from Albania and blocked from returning, whereas Macedonia was unlikely to tolerate the permanent residence of large numbers of Albanians from Kosovo.)

Whatever the explanation, the practice of "identity cleansing" was clearly not a random initiative by Serbian officials on the border. After the war, piles of license plates and burned documents were discovered by the border crossings into Albania and elsewhere in Kosovo.

Each one of these four factors (the timing of arrivals, the means of departure, the use of terror and the practice of "identity cleansing") strongly suggest that the flight of some 860,000 Albanians from Kosovo in twelve weeks adds up to systematic forced expulsion. Taken together, the evidence is overwhelming. The prosecutor of the ICTY was in no doubt when her office prepared the indictment for Slobodan Milosevic, Milan Milutinovic, Nikola Sainovic, Dragoljub Ojdanic, and Vlajko Stojiljkovic (see Forces of the Conflict). The first of the charges reads:

The forces of the FRY and Serbia have, in a systematic manner, forcibly expelled and internally displaced hundreds of thousands of Kosovo Albanians from their homes across the entire province of Kosovo. To facilitate these expulsions and displacements, the forces of the FRY and Serbia have intentionally created an atmosphere of fear and oppression through the use of force, threats of force, and acts of violence.70

Arbitrary Arrests and Detentions

Throughout the armed conflict in 1998 and early 1999, arbitrary arrests and detentions of Kosovar Albanians were commonplace. Physical abuse and torture of detainees was widespread.71 The practice intensified during the period March-June 1999. Thousands of Kosovar Albanians were detained during the NATO bombing. Very often, men were separated from women and held in makeshift detention centers, such as schools or factories, for a period of days, during which time they were beaten and interrogated about the KLA. Some men were held for longer periods in the prisons at Smrekovnica, Lipljan, Istok (Dubrava), or Pristina. Most of the men in detention at the end of the war-between 1,000 and 2,000-were transferred out of Kosovo to prisons in Serbia proper; as of March 2001, more than 400 Kosovar Albanians were still being held.

As an example, more than 300 men and women were taken into detention from the streets and private homes in Djakovica between May 7 and May 11 after fighting in the Cabrat neighborhood between government forces and the KLA (see Djakovica Municipality). Women were held temporarily in the Gorenje Elektromotor factory and then released. The men were held in an unknown location on the edge of the city along the Djakovica-Pec road.

Approximately half of the men were released after six days but the rest-an estimated 150 people-were transferred to the jails in Pec, Lipljan, and then Dubrava (prior to the NATO bombing of that prison). Most, if not all, of the detainees were transferred to prisons inside Serbia just before NATO troops entered Kosovo on June 12. On May 22, 2000, a court in Nis sentenced 143 men from Djakovica to a total of 1,632 years in prison for acts of terrorism. The Serbian Supreme Court released them on April 23, 2001.

More than 300 ethnic Albanians were held in Prizren's prison beginning in April. One man, who spent two months in detention, spoke about the treatment he and others received:

I went straight to the prison; they never brought me to the police station. There they tortured me for breakfast, lunch, and dinner. The first thing they did to me when I arrived was beat me using rubber police batons. They hit me on the palms of my hands and in my groin. There were more than 300 Albanians there. Every time they served us food, they beat us. During my first four days, they beat me every day in the hall. Other inmates got the same treatment.72

In the town of Glogovac, police raided neighborhoods and detained large groups of adult men in the local police station for one week beginning April 22 (see Glogovac Municipality). Almost all of the men were beaten in front of their homes or on the way to the station, and some were forced to sing Serbian nationalist songs. Most of the detainees were questioned about the KLA and then released after no more than one day in custody. One month later, hundreds of Glogovac men were detained and held in the basements of local shops. The majority of the men were released after three days, but only after interrogations and beatings with sticks, shovel handles, and metal bars. Approximately ninety men were transferred to Lipljan.

In some villages, women were held in makeshift detention facilities, such as barns or abandoned homes, where they were sexually abused or raped (see section on rape).73 A few women were reported to have been held in Kosovo's prisons, such as the doctor Flora Brovina, who was transferred to Serbia proper after the war, sentenced to twelve years in prison, but released in November 2000 (see Abuses After June 12, 1999).

More than 2,000 Kosovar Albanian men were detained in Smrekovnica prison near Mitrovica in May (see section on Smrekovnica prison in Vucitrn Municipality). Human Rights Watch interviewed more than thirty of these men; all of them reported regular beatings by police during their detention, especially during interrogations about the KLA. Signs of physical abuse, such as black eyes, severe bruises, and skin abrasions were visible on their bodies when they were interviewed as refugees in Albania. Most of the men were forced to sign confessions that they were engaged in terrorist activities before being released.

Human Rights Watch also spoke with former detainees from Lipljan prison, who testified to the beatings and indecent conditions at the prison, including a cordon of policemen and prison guards who beat new arrivals. One former prison from Glogovac, who was in Lipljan for fifteen days before being transferred to Sremska Mitrovica prison on June 10, said:

At first when we got out of the bus [at Lipljan prison] we had to walk thirty meters through two columns of police, and they beat us with bars and sticks. Inside they told us to disrobe. And they beat us again. They took us to another room with about 500 prisoners-in a sports hall. We all stayed together.74

Another detention facility was Dubrava penitentiary near Istok-Kosovo's largest prison. Between 900 and 1,100 prisoners were being held there when the NATO bombing began, including approximately thirty ethnic Serbs, but additional prisoners were transferred to Dubrava after the bombing had begun (including the 150 men from Djakovica), some of them from prisons in Serbia proper.

NATO bombed the prison on May 19 and May 21, killing at least nineteen ethnic Albanian prisoners (see section on Dubrava prison in Istok Municipality). On May 22, Serbian security forces lined up the approximately 1,000 prisoners in the courtyard and fired on them with snipers, machine guns, and grenades from the prison walls and guard towers, killing at least seventy people. At least another twelve prisoners were killed over the next twenty-four hours as prison guards, special police, and possibly paramilitaries attacked prisoners who were hiding in the prison's undestroyed buildings, basements, and sewers. The injured were taken away in trucks, while the remaining prisoners were transported to Lipljan prison, where they were beaten. On June 10, they were transferred to prisons in Serbia proper.

On June 10, two days before NATO's entry into Kosovo, an estimated 2,000 ethnic Albanian prisoners were transferred from Kosovo to prisons inside Serbia proper. More than 1,400 of these people were released at different times in 1999, 2000, and 2001, some of them under a Yugoslav amnesty law passed in February 2001. As of March 2001, approximately 400 Kosovo Albanians were known to be in Serbian prisons. (See Abuses After June 12, 1999.)

Destruction of Civilian Property and Mosques

Between March and June 1999, Yugoslav soldiers and Serbian police and paramilitaries destroyed thousands of Albanian homes throughout Kosovo with the use of artillery, bulldozers, explosives, and arson. There was also widespread looting. In some areas, especially those with close ties to the KLA, entire villages were destroyed. A number of towns, including Pec and Glogovac were also badly damaged. Albanian-owned shops and businesses, schools, and mosques were also targeted for destruction and were frequently looted. In addition, there was widespread contamination of water wells in the province (see following section).

These actions were the continuation of policies carried out by security forces in Kosovo during 1998 and early 1999, notably in the Drenica region and Orahovac municipality (particularly in the town of Malisevo) and Decani municipality.75 A UNHCR shelter survey released in November 1998 demonstrates the widespread nature of the destruction during 1998.76 The survey assessed 285 villages, of which 210 had been affected by the conflict. In the 210 affected villages with an estimated pre-conflict population of 350,000 persons, 28 percent of the homes-9,809 out of a total of 35,185 homes-had been completely destroyed. Another 15 percent of the homes (5,112 homes) had severe damage, while an additional 6,017 homes sustained moderate to minor damage, leaving only 40 percent of the homes in the affected regions undamaged.

Much of the remaining housing stock in Kosovo was destroyed in 1999. According to a November 1999 UNHCR survey, almost 40 percent of all residential houses in Kosovo were heavily damaged (categories III and IV) or completely destroyed (category V).77 Out of a total of 237,842 houses, 45,768 are heavily damaged and 46,414 are destroyed. Municipalities with strong ties to the KLA were disproportionately affected (probably in part because they began to be attacked in 1998): almost half the 12,887 houses in Orahovac municipality were heavily damaged (4,334 houses) or completely destroyed (1,943 houses); in Suva Reka, 4,552 homes were heavily damaged and 2,018 destroyed, or more than 55 percent of the total number (11,622). In the Drenica region, out of a total housing stock of 17,340 units, 7,155 were heavily damaged and 6,209 completely destroyed, or 77 percent of the total. The city of Pec was also hit especially hard, with more than 80 percent of the city's 5,280 houses heavily damaged (1,590) or destroyed (2,774).

Schools and mosques were similarly affected. According to a United Nations damage assessment of 649 schools in Kosovo, more than a fifth of the schools surveyed were heavily damaged and more than 60 percent were completely destroyed.78 After June 1999, Human Rights Watch observed damaged and destroyed mosques in Djakovica, Pec, Istok, and Cirez (Srbica municipality). Field visits by Human Rights Watch over the summer of 1999 also revealed extensive damage to shops and businesses in Suva Reka, Urosevac, Pristina, Pec, Glogovac, and Djakovica. An August 1999 report by Physicians for Human Rights documented 155 destroyed mosques throughout Kosovo, based on refugee accounts.79

Contamination of Water Wells

One of the more blatant forms of civilian property destruction in Kosovo during 1998 and 1999 was the widespread practice of water well contamination, which is forbidden by the laws of war.80 Throughout Kosovo, Serbian and Yugoslav forces deliberately rendered water wells unusable by disposing of chemicals, dead animals, and even human corpses into the water. According to the International Committee of the Red Cross (ICRC), which ran a water-sanitizing operation in Kosovo, "Of the 20,000 wells in Kosovo, over half are believed to have been contaminated with animal or human remains or with rubbish, or have simply grown stagnant through lack of use." Between January and September 1999, the ICRC cleaned over 1,700 wells.81

Human Rights Watch documented a number of villages in Kosovo where murder victims were dumped into a local well. In one village in Suva Reka-which is not named because of the rape that occurred there-eleven men were killed in April 1999 and thrown into the well (see Suva Reka Municipality). As noted in the above discussion of the death toll, at least four of the estimated forty-seven members of the Muqolli family, executed in the village of Poklek in April 1999, were thrown into a well in the yard of the family compound. In another village in Drenica region (that will also remain nameless to protect the identity of the victims), eight women were raped, shot, and then thrown into a well.82 Human Rights Watch has also received reports that human remains were found in water wells in the villages of Donji Streoc (Strellc i Poshtem) and Dubovik in Decani municipality, the village of Damjane (Dehje) in Djakovica municipality, and Studenica (Studenice) in Istok municipality. The two wells outside the village of Kotlina, into which twenty-three men were thrown together with hand-grenades in April 1999, were not used by the villagers for water and the decision to put the men into the wells appears to have been an attempt to conceal their murders rather than to contaminate the water.

Secondary sources also reported that wells had been contaminated, both prior to and during the NATO bombing. According to an article in The Washington Post on December 10, 1998, at least fifty-eight villages had informed foreign aid organizations that their wells contained "dead dogs, chickens, horses, garbage, fuel oil, flour, detergent, paint and other contaminants." In one village, Ovcarevo, 70 percent of the wells "might be contaminated."83 After the NATO bombing, The Los Angeles Times cited UNHCR statistics on the Djakovica area indicating that the wells in thirty-nine of forty-four villages were contaminated with "either human or animal bodies."84 According to the Council for the Defense of Human Rights and Freedoms in Pec, human remains were found in wells in the villages of Donja Luka (Lluke e Eperme), Banja (Banje), and Kosuric (Kosuriq).85 The Italian newspaper Corriere Della Sera also reported on the contamination of the Donja Luka well, in which a villager found thirteen relatives and neighbors, aged twelve to seventy.86

Robbery and Extortion

The killing and mass expulsion in Kosovo in the spring of 1999 was often accompanied by robbery and extortion. Typically, Albanians were robbed immediately prior to or during their expulsion from Kosovo. Those who refused were threatened with death, or their children were snatched and threatened until they complied. Serbian paramilitaries or persons in unknown uniform were most often implicated in the crimes, although there are also reports of the involvement of Serbian police. Incidents of robbery were not confined to any particular area: there are notable examples in regions with significant KLA activity and widespread killings, such as Glogovac, Suva Reka, and Pec, and in regions that experienced fewer killings and relatively light destruction, such as Lipljan, Pristina, and Gnjilan. As in Bosnia and Herzegovina, robbery and extortion may simply have been motivation and reward for Serbian paramilitaries, although the widespread nature of the abuses and the involvement of the police in at least some cases suggests complicity by, and possibly profit to, the state.

One of the most egregious cases of mass robbery occurred in the village of Belanica in Suva Reka municipality (see Suva Reka Municipality). On March 31, Serbian police and paramilitaries entered the village and rounded up residents and the large number of internally displaced persons from the surrounding area. Serbian paramilitaries and police then repeatedly demanded money, threatening those who refused or their children with death. They made good on their threats: at least a dozen Albanians were killed. A woman who was in the village at the time told Human Rights Watch:

At one point, the police came up to a man on the tractor in front of me. They said, "Give us money!" He didn't have anything for them. He was from Ostrozuk village. So they pulled him off the tractor and killed him. When he didn't give them anything they [four policemen] pulled him off the tractor by his arms and legs. They brought him around to the back of a house, then I couldn't see him anymore. I heard shooting and I could see one of the policemen aiming his gun and firing. The man didn't come back to his _tractor.87

In the weeks leading up to the mass expulsion from Glogovac at the start of May 1999, the town's residents and displaced persons suffered persistent visits to their homes by Serbian paramilitaries and police demanding money and valuables (see Drenica Region). In some areas, demands for money by paramilitaries were so frequent that residents took the unusual step of going to the police on April 23 to request protection. Protection was provided only intermittently and did little to curtail criminal activity, suggesting a degree of official complicity.

The robberies followed a similar pattern: one or two paramilitaries or police officers would break in the door of a private apartment, sometimes wearing masks, but always carrying automatic rifles. The families were physically threatened until they handed over everything of value. A fifty-nine-year-old man from Glogovac, A.H., described what happened when men with "green uniforms and red bandanas on their arms" came to the four-house compound he shared with his three brothers in early May:

Two days before we left, at around 9:00 or 10:00 a.m., they [the police] came into the house and searched us . . . They pointed their guns at us . . . They asked me for money . . . [then] they forced me to strip to my underwear - looking for money. One of them said, "If I find any money on your body, I'm going to shoot you . . ." They took rings and gold from the women . . . The next day . . . they took two radios from my brother and a small TV.88

During some of the robberies, paramilitaries and police reportedly threatened children with knives and automatic guns in order to extort money from their parents. According to H.M., a forty-six-year-old man, from Glogovac: "A week before we left [paramilitaries] started to take very strong action to take money. They would take your daughter and say, `Give me money or I won't let her go.'"89 Another man from Glogovac in his late fifties said that "paramilitaries came, they took children, held a knife against their throats [and threatened to kill them] unless they were given money."90

Areas that escaped multiple killings and widespread destruction were nonetheless targeted for robbery and extortion. Residents of Pristina reported frequent robberies as they were forced from the city at the end of March. The municipality of Gnjilane, which was among the least affected by the war, also experienced robbery and extortion as its population was forced to flee toward Macedonia. The experiences of the residents of Malesevo village are typical of the pattern in Gnjilane. After green-uniformed paramilitaries with "a white eagle with four C's"91 insignia entered their village on the morning of April 16, Malesevo residents were given two hours to leave. Two villagers were shot in a field as the rest of the village stood by. The villagers were beaten, robbed of cars, money, and jewelry, and threatened with death before they were forced to walk to nearby Gnjilane. Later, on the road between Gnjilane and Urosevac, the villagers were repeatedly robbed by bands of paramilitaries. One villager said that the column had been stopped nine times between Gnjilane and the town of Klokot (Kllokot), approximately ten kilometers by road. Each time the villagers were stopped they were slapped, beaten, and threatened with death if they did not hand over all their foreign currency, jewelry, and other valuables. Several witnesses indicated that paramilitaries had threatened to detonate hand grenades if their demands were not met. According to one witness: "They took hand grenades and were threatening that they would throw them in the midst of the children if they didn't get anything."92

Detentions and Compulsory Labor

There is clear evidence that between March and June 1999, Yugoslav and Serbian security forces detained adult males and compelled them to dig trenches, clear bunkers, and perform other manual work. The majority of the work appears related to the Yugoslav Army's strategic objectives. The most notable incidents occurred in Glogovac municipality during May and June 1999. Human Rights Watch also documented the compulsory labor of detainees in Prizren municipality during April 1999, where on at least two occasions, Serbian police and Yugoslav soldiers rounded up men in Prizren town and transported them to the Albanian border where they were forced to serve on trench-digging brigades (see Prizren Municipality). There are also reports of work brigades in Orahovac, Suva Reka, and Djakovica, where Roma were allegedly compelled to dig trenches on the border with Albania during the month of April 1999.93

Human Rights Watch received detailed accounts of compulsory labor from the group of prisoners who were detained at the mosque in Cirez in Drenica on April 31, 1999. While many of the detainees were executed at the Shavarina mine near Cikatovo on May 1, 1999, the remainder were transferred to Glogovac (see Drenica Region). Around seventy-six of the survivors were taken by Serbian police from Glogovac to the villages of Krajkovo (Krajkove), Vukovce (Vukofci), and Poturk on May 5 and 6 and handed over to Serbian soldiers. According to seven witnesses interviewed by Human Rights Watch, the men were then forced to work for approximately six weeks until the withdrawal of Serbian security forces from Kosovo in mid-June. Tasks included digging trenches and bunkers. Witnesses reported incidents of beating and torture of prisoners at the hands of the soldiers, although some noted that they were fed and received better treatment than they had at the hands of Serbian police and paramilitaries in Cirez and Glogovac.

Human Shields

Between March and June 1999, Yugoslav and Serbian security forces compelled some Albanian civilians to remain close to them or situated them between Serbian positions and those of the KLA. Both strategies were designed to create a human shield to protect Yugoslav and Serbian forces from attack from NATO aircraft or the KLA. The use of such "human shields" is prohibited by Article 28 of the Fourth Geneva Convention, which states that "The presence of a protected person may not be used to render certain points or areas immune from military operations."

In some cases, the use of human shields was direct, with Albanian civilians compelled to march alongside Yugoslav soldiers and Serbian police in order to protect them from attack by NATO aircraft or the KLA. Human Rights Watch interviewed witnesses from Suva Reka, Klina, and Lipljan who described being held as human shields. Witnesses reported to the OSCE that they were detained as human shields in Djeneral Jankovic (on the Macedonian border), in Djakovica municipality, in Pristina, and on the road between Cirez and Stutica.94 There are also allegations that the displaced Albanian civilians killed in the Korisa woods by NATO bombs had been compelled by Serbian forces to remain there as "human shields" but the evidence is inconclusive (see The NATO Air Campaign).95

There is also evidence of the indirect use of human shields, through efforts to create barriers of civilians between Serbian security forces and the KLA, thereby limiting the ability of the KLA to attack, or defend from attacks by, Serbian positions. Most of the reports are from the Drenica region. The case of Staro Cikatovo village is a notable example (see Drenica Region). One witness from the village told Human Rights Watch, "We were between the KLA and Feronikel. [Serbian forces] started grenading from Feronikel to attack [KLA] soldiers."96 The OSCE reports a similar incident in the village of Trnavce (Trnoc) in Srbica municipality, as well as in Belanica village in Suva Reka municipality.97 A witness from Malisevo (Orahovac municipality) interviewed by Physicians for Human Rights reported being in a group of 500 civilians who were detained close to the Albanian border and used as human shields by Yugoslav soldiers as they attacked a nearby KLA position.98


Throughout 1998 and the first six months of 1999, Serbian and Yugoslav forces placed an estimated 50,000 anti-personnel and anti-tank mines in Kosovo, especially along the borders with Macedonia and Albania.99 The KLA also placed mines in areas under its control during this period. In addition, NATO's use of cluster bombs during at least six weeks of the air war resulted in some civilian deaths and unexploded ordinances (UXO) scattered in Kosovo and elsewhere in the Federal Republic of Yugoslavia after the war.

Mines and UXO represent the deadliest legacy of the conflict in Kosovo. According to the U.S.-based Vietnam Veterans' Foundation, from the end of the war to mid-November 2000, 103 people were killed and 394 wounded in accidents with mines and unexploded ordinance.100 This differs only slightly from a July 2000 report of the U.N. secretary general, which said that 101 people had been killed and 395 injured in mine or unexploded ordnance incidents between June 12, 1999, and July 2000.101 The U.N. Mine Action Coordination Center (MACC) established in Kosovo after the war, reported that by mid-July 2000, 1.1 million square meters of land had been demined or cleared of unexploded ordnance.102 The secretary general's report also noted that teams coordinated by MACC had cleared 3,405 anti-personnel mines, 3,768 anti-tank mines, 3,066 cluster bombs, and 9,327 items of unexploded ordnance. In addition, KFOR had cleared over 16,000 homes, 1,165 schools, and almost 2,000 kilometers of roads during the same period.103

The Yugoslav government, one of the world's largest producers of anti-personnel mines, is not a signatory to the 1997 treaty banning the use of landmines, but Yugoslav officials have stated that landmines are used only for the purposes of training.104 In fact, landmines were used extensively by Yugoslav forces during the 1998-99 conflict as a tactical measure against the KLA and NATO. According to the International Campaign to Ban Landmines 2000 report, the Yugoslav Army primarily placed barrier minefields along the southern border with Macedonia. Serbian Ministry of Internal Affairs forces and paramilitaries laid anti-personnel mines in and around civilian population centers.105 According to a report by the HALO Trust, a non-profit demining group, "[M]any villages are afflicted with random mines laid with the sole aim of causing civilian casualties and thereby discouraging the return of refugees."106 According to a U.S. Agency for International Development report, more than 900 schools needed mine clearance.107

In the Military Technical Agreement signed between NATO and the Yugoslav government on June 9, 1999, the Yugoslav authorities agreed to "mark and clear minefields, booby traps and obstacles."108 The Yugoslav Army informed NATO of 616 mined areas. Mines placed by Serbian paramilitaries, however, were generally unrecorded and are therefore difficult to locate.109

The same is true for many of the mines planted by the KLA. The KLA is known to have placed mostly anti-tank mines but also some anti-personnel mines around regional bases, headquarters, and safe houses. In late September 1998, a Canadian armored car from KDOM and a vehicle from the ICRC both hit anti-tank mines laid by the KLA in central Drenica. A Kosovo Albanian doctor, Dr. Shpetim Robaj, was killed and three ICRC medical workers were injured.

Cluster bombs dropped by NATO forces during the air war also pose a continued risk to civilians in post-war Kosovo. Both U.S. and British forces have acknowledged the use of cluster bombs in the bombing campaign.110 The White House issued a directive to the Pentagon in May 1999 to restrict cluster bomb use (at least by U.S. forces) after fourteen civilians were killed and twenty-eight were injured in a May 7 cluster bomb attack on an airfield in Nis (see The NATO Air Campaign).111

According to the International Campaign to Ban Landmines, "[T]he U.S dropped 1,100 cluster bombs of the type CBU-87/B, each containing 202 BLU-97/B bomblets and the UK dropped 500 RBL/755 cluster bombs, each containing 147 Bl-755 bomblets."112 Because of the high failure rate of the submunitions inside cluster bombs-estimated conservatively at 5 percent113- these "bomblets" in effect become anti-personnel landmines.

1 According to UNHCR statistics, the total number of refugees from Kosovo on June 9 was 862,979. This figure excludes those who had sought asylum in Europe prior to March 1999. No precise figures exist for the total population of internally displaced in Kosovo between March and June 1999 but most estimates range between 500,000 and 600,000, which includes the more than 200,000 persons internally displaced prior to March.
2 The expulsion of virtually all foreign journalists from Pristina on March 25 completed the removal of all foreign witnesses from Kosovo.
3 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, Part I, p. 37.
4 See the following press accounts: "[T]he death toll from Kosovo's war was about 2,000 at the end of last year [1999], but that figure is substantially higher now," Anne Thompson, "A worsening terror: Disappearances Grow in Kosovo," Associated Press, March 18,1999; ". . . more than 1500 people have been killed," Charles Trueheart, "Kosovo Delegation Vows Anew to Sign Peace Agreement; West to Reapply Pressure on Belgrade," Washington Post, March 16, 1999; ". . . more than 2,000 deaths," Kurt Schork (Reuters), "A Year Ago, Serb Attack Kicked Off Kosovo War," Seattle Times, March 5, 1999. See also, The Kosovo Report, The Independent International Commission on Kosovo, October 2000.
5 UNHCR Press Briefing Note: Kosovo, Tuesday, February 2, 1999.
6 According to a November 1998 survey of 285 villages in Kosovo carried out jointly by UNHCR and a group of humanitarian NGOs, 210 villages had been affected by the fighting in 1998. Twenty-eight percent of the homes in those villages (9,809) had been completely destroyed, and 15 percent (5,112) of the homes had severe damage. [UNHCR Pristina,"IDP/Shelter Survey Kosovo: Joint Assessment in 20 Municipalities," November 12, 1998.]
7 Out of an estimated Kosovo population of 1.8 million, 850,000 Kosovars were refugees, and as many as 600,000 were internally displaced. Given that approximately 200,000 of the total population were Serbs, Roma, and other minorities, the percentage of displaced Albanians from Kosovo may have been as high as 90 percent.
8 UNHCR contingency plans prepared by the UNHCR Special Envoy for the Former Yugoslavia prior to March 24 put the maximum number of refugees from Kosovo that could be expected as a result of intensified hostilities in the province at 100,000. "The Kosovo refugee crisis: an independent evaluation of UNHCR's emergency preparedness and response," UNHCR Evaluation and Policy Analysis Unit, February 2000.
9 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, Part I, pp. 21-30.
10 Ibid.
11 For a Seselj statement on forced expulsions published in The Greater Serbia Journal on October 14, 1995, see:, (accessed March 2001).
12 Human Rights Watch found few cases of "identity cleansing" among refugees arriving in Macedonia. This may be due to the weaker base of support for the KLA in eastern Kosovo (the point of origin for most refugees arriving in Macedonia). Another possibility is that Macedonia, unlike Albania, would not permit such a large influx of Kosovo Albanians to remain indefinitely.
13 Bolstered by four years of a United Nations preventive deployment of peacekeepers, Macedonia had avoided the civil conflict that engulfed the other republics of the Socialist Federal Republic of Yugoslavia. The election of a multi-ethnic government coalition had mitigated tensions between the large ethnic Albanian community, whose experience of persistent discrimination had strengthened calls among many of its members for greater autonomy, and the larger ethnic Macedonian population, whose fears of a greater Albania were inflamed by such calls.

For more on human rights in Macedonia, see Human Rights Watch, "Police Violence in Macedonia," April 1998, and Human Rights Watch, A Threat to Stability: Human Rights in Macedonia (New York: Human Rights Watch, 1996).
    14 The Macedonian government periodically closed its borders to Kosovo refugees until the international community agreed to transfer out of the country some of those already present in Macedonia. For more on the Macedonian government's treatment of refugees during the war, see a Human Rights Watch statement, "Macedonia Must Keep Border Open To Refugees," April 1, 1999.
    15 After the end of the war, some western politicians even claimed that the return of refugees was the original objective of the NATO action-a temporal impossibility, further contradicted by NATO's own predictions about possible refugee flows that might follow the start of the bombing.
    16 Human Rights Watch interviews, Bogovine, Macedonia, April 5, 1999; Neprosteno, Macedonia, April 10-11, 1999; Kukes, Albania, April 17 and April 26, 1999; Kukes, Albania, May 12-13 and May 22, 1999; Kukes, Albania, June 6, 1999.
    17 Prior to March 1999, more than 90 percent of the population of Leposavic and approximately 75 percent of the populations of Zubin Potok and Zvecan were Serbs.
    18 Human Rights Watch interviews, Bogovine, Macedonia, April 4, 1999.
    19 Human Rights Watch interviews, Kukes, Albania: April 16-19, 1999; April 23, 1999; April 26, 1999; May 10, 1999; May 11, 1999; May 13, 1999; May 22, 1999; June 2, 1999 and June 7, 1999.
    20 Many of them were displaced again during the forcible expulsion of large sections of Pristina in March and April 1999. (Human Rights Watch interviews, Dzepciste, Macedonia, April 3, 1999).
    21 Human Rights Watch interviews, Podujevo town, Dobratin village (Podujevo municipality), July 13, 1999; Kukes, Albania, May 22, 1999.
    22 For more information on events in Kosovo Polje and Obilic municipalities during 1998, see: OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, pp. 235-239 and pp. 268-273. Some details on the KLA's brief capture of the Belacevac mine, and the nine Serbs who went missing, can be found in a Human Rights Watch report, Humanitarian Law Violations in Kosovo, October 1998.
    23 Human Rights Watch interviews, Bogovine, Macedonia, April 4, 1999; Kukes, Albania, April 4 and April 14, 1999; Kukes, Albania, June 8, 1999.
    24 Human Rights Watch interviews, Bogovine, Macedonia, March 31, 1999 and April 5, 1999; Kukes, Albania, April 29, 1999 and May 10, 1999.
    25 For more information on Ade see: Human Rights Watch, Kosovo Human Rights Flash #32, April 29, 1999.
    26 Human Rights Watch interviews, Morina, Albania, April 28, 1999; Kukes, Albania, April 29, 1999; Tetovo, Macedonia, April 30, 1999.
    27 Human Rights Watch interviews, Krume, Albania, April 2, 1999; Kukes, Albania, April 8 and 9, 1999.
    28 Human Rights Watch interviews, Morina, Albania, April 4, 1999; Kukes, Albania, April 8, 1999; Morina, Albania, April 28, 1999.
    29 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, pp. 216-226.
    30 Human Rights Watch interviews, Neprosteno, Macedonia, April 8 and 11, 1999; Zelino, Macedonia, April 12, 1999; Orasje, Macedonia, April 14, 1999; Senekos, Macedonia, April 16 and 17, 1999; Gostivar, Macedonia, April 18, 1999.
    31 Human Rights Watch interviews, Orasje, Macedonia, April 12 and 14, 1999; Kotlina, Kosovo, August 8, 1999. See also, Amnesty International, "FRY: Killings in the Kacanik Area," April 9, 1999.
    32 Human Rights Watch interviews, Senekos, Macedonia, April 16, 1999; Gostivar, Macedonia, April 18, 1999; Kukes, Albania, April 18, 1999; Kukes, Albania, May 14, 1999.
    33 Human Rights Watch interviews, Neprosteno, Macedonia, April 8, 1999; April 10-11, 1999.
    34 "Yugoslavs are Crying `Crocodile Tears': Cohen - 100,000 Kosovars May be Dead, Says Defense Secretary," Toronto Star, May 17, 1999; David E. Rosenbaum, "Crisis in the Balkans: the Dead: U.S. Official Calls Tallies of Kosovo Slain Too Low," New York Times, April 19, 1999.
    35 U.S. Department of State, "Erasing History: Ethnic Cleansing in Kosovo," May 1999; Transcript of Press Conference by Jamie Shea and Brigadier General Giuseppe Marani, NATO Headquarters, April 17, 1999.
    36 "UN Sets Kosovo Dead at 11,000," Reuters, August 3, 1999, and "Kouchner's Spokeswoman Comments on Kosovo Death Figures," Agence France Press, August 3, 1999.
    37 See, for example: Charles A. Radin and Louise D. Palmer, "Number of missing Kosovars is challenged," The Boston Globe, April 21, 1999.
    38 For transcripts of the program, plus two other reports on war crimes in Kosovo, see:, (accessed March 2001).
    39 "Milosevic Tried to Cover up Kosovo Crimes: Official," Agence France Presse, May 26, 2001, and "A Dark Secret Comes to Light in Serbia," by Carlotta Gall, New York Times, June 1, 2001.
    40 Beta News, July 17, 2001, and Madame Carla Del Ponte, Prosecutor, ICTY, "Remarks to the Security Council," November 10, 1999.
    41 ICRC Statement, "Persons Unaccounted for in Connection with the Kosovo Crisis," April 10, 2001. Prior to this statement, ICRC had said that 3,368 persons from Kosovo remained missing as of June 27, 2000. The majority of the missing are Kosovo Albanians, but also include 400 Serbs, one hundred Roma and persons from other minorities. Seventy-four percent disappeared between March and June 1999. ICRC, "Update 00/01 on ICRC activities in the Federal Republic of Yugoslavia," June 27, 2000.
    42 Madame Carla Del Ponte, Prosecutor, ICTY, "Remarks to the Security Council," November 10, 1999.
    43 Address to the Security Council by Madame Carla Del Ponte, prosecutor, International Criminal Tribunals for the former Yugoslavia and Rwanda, November 21, 2000, New York.
    44 Human Rights Watch interview with N.B., Studencane, Kosovo, August 29, 1999.
    45 Paul B Spiegel and Peter Salama, "War and mortality in Kosovo, 1998-99: an epidemiological testimony," The Lancet, Vol. 355, No. 9222, June 24, 2000.
    46 Political Killings in Kosova/Kosovo, Central and East European Law Initiative of the American Bar Association and the American Association for the Advancement of Science, Washington D.C., October 2000.
    47 Other known lawyers killed during the war were Urim Rexha from Djakovica, Mehdi Elshani from Suva Reka, and Ismet Gashi from Prizren.
    48 Human Rights Watch interview with Nekibe Kelmendi, Pristina, Kosovo, July 25, 1999.
    49 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, Part I, p. 236.
    50 May 7, 1999, Announcement, MUP website,\mup.nsf/_pages/index, (accessed March 2001).
    51 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, Part I, pp. 151-152.
    52 For a tribute to Mehmeti and Dervishi, see Robert Elsie, "Gone but not Forgotten," Guardian, April 3, 1999.
    53 The findings of this research are set out in full in a separate Human Rights Watch report: "Kosovo: Rape as a Weapon of Ethnic Cleansing," A Human Rights Watch Report, vol. 12, no. 3, March 2000.
    54 Human Rights Watch received credible reports of rape from the Center for the Protection of Women and Children, based in Pristina (Prishtina) (twenty-nine cases); the Albanian Counseling Center for Women and Girls, an NGO in Albania (twenty-eight cases); the Yugoslavia-based Humanitarian Law Center (four cases) and; the Kosovo-based Council for the Defense of Human Rights and Freedoms (four cases). Médécins Sans Frontières (MSF), with offices in Kosovo before and after the war, reported four cases, and other medical personnel working in Kosovo and Albania confirmed an additional eight cases. Physicians for Human Rights interviewed four victims of sexual violence, and Amnesty International documented another three cases of rape, although two were also counted by Human Rights Watch.
    55 Human Rights Watch interview, R.G., Kukes Refugee Camp, Albania, June 5, 1999.
    56 Human Rights Watch interview, P.J., Kukes Refugee Camp, Albania, June 5, 1999.
    57 UNHCR Press Briefing Note: Kosovo, Tuesday, February 2, 1999.
    58 UNHCR Press Briefing Note: Kosovo, April 13, 1999.
    59 UNHCR Press Briefing Note: Kosovo, May 13, 1999.
    60 Statistic from: "The Kosovo refugee crisis: an independent evaluation of UNHCR's emergency preparedness and response," UNHCR Evaluation and Policy Analysis Unit, February 2000. This figure excludes those who had sought asylum in Europe prior to March 1999.
    61 No precise figures exist for the total population of internally displaced in Kosovo between March and June 1999 but most estimates range between 500,000 and 600,000.
    62 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told. pp. 98-99.
    63 The American Association for the Advancement of Science, East-West Management Institute and the Institute for Legal and Policy Studies, Policy or Panic: The Flight of Ethnic Albanians from Kosovo, March-May 1999, May 2000. The report is available at:, (accessed March 2001).
    64 Human Rights Watch interview with M.B., Skopje, Macedonia, April 2, 1999.
    65 Human Rights Watch interview with X.P., Skopje, Macedonia, April 2, 1999.
    66 Human Rights Watch interview with N.J., Tetovo, Macedonia, March 31, 1999.
    67 Human Rights Watch interview with S.K., Skopje, Macedonia, April 2, 1999.
    68 See Human Rights Watch, Kosovo Human Rights Flash # 16, "Violent Ethnic Cleansing in Djakovica," April 3, 1999.
    69 Human Rights Watch interview, name unknown, Krume, Albania, April 2, 1999.
    70 International Criminal Tribunal for the former Yugoslavia, "Prosecutor of the Tribunal against Slobodan Milosevic and others: Indictment," June 1999.
    71 See Human Rights Watch, "Detentions and Abuse in Kosovo," A Human Rights Watch Report, vol. 10, no. 10, December 1998.
    72 Human Rights Watch interview with A.S. Prizren, Kosovo, June 13, 1999.
    73 See Human Rights Watch, "Kosovo: Rape as a Weapon of `Ethnic Cleansing'."
    74 Human Rights Watch interview with R.N., Glogovac, Kosovo, November 5, 1999.
    75 For more information on destruction during 1998, see Human Rights Watch, A Week of Terror in Drenica, February 1999. pp. 57-62.
    76 UNHCR Pristina, "IDP/Shelter Survey Kosovo: Joint Assessment in 20 Municipalities," November 12, 1998.
    77 UNHCR GIS Unit, Pristina, Kosovo, "UNHCR Shelter Verification: Agency Coverage," November 9, 1999.
    78 Status Report, United Nations Interim Administration Mission in Kosovo, July 14, 2000.
    79 Physicians for Human Rights, War Crimes in Kosovo: A Population-Based Assessment of Human Rights Violations Against Kosovar Albanians, August 1999.
    80 Protocol II, Article 14, of the Geneva Conventions states: "It is therefore prohibited to attack, destroy, remove or render useless, for that purpose, objects indispensable to the survival of the civilian population, such as foodstuffs, agricultural areas for the production of foodstuffs, crops, livestock, drinking water installations and supplies and irrigation works."
    81 "Balkan Crisis: Cleaning Wells in Kosovo," ICRC News 99/37, September 8, 1999.
    82 For more information on the case, see: Human Rights Watch, "Kosovo: Rape as a Weapon of `Ethnic Cleansing'."
    83 R. Jeffrey Smith, "Carcasses Dumped in Wells as Serb Crackdown Ended," Washington Post, December 10, 1998.
    84 Valerie Reitman, "Kosovo Wells Emerging as Mass Graves," Los Angeles Times, August 10, 1999.
    85 Human Rights Watch interview with Dr. Asllani, Pec, Kosovo, July 23, 1999.
    86 Corriere Della Sera, July 6, 1999.
    87 Human Rights Watch interview, Kukes, Albania, April 4, 1999.
    88 Human Rights Watch interview with A.H., Cegrane, Macedonia, May 15, 1999.
    89 Human Rights Watch interview with H.M., Stenkovac II refugee camp, Macedonia, May 10, 1999.
    90 Human Rights Watch interview with I.X., Cegrane, Macedonia, May 15, 1999.
    91 The four C's insignia is a Serbian nationalist symbol, comprising a cross and four Cyrillic S's. It is derived from the slogan "Only Unity Saves the Serbs" (Samo Sloga Srbina Spasava).
    92 Human Rights Watch interview with H.S., Neprosteno, Macedonia, April 22, 1999.
    93 International Crisis Group, Reality Demands: Documenting Violations of International Humanitarian Law in Kosovo 1999 (Brussels: International Crisis Group, 2000).
    94 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, pp. 94-96.
    95 See also, Ian Fisher, "Refugees `Human Shield' Witnesses Say They Were Held in Targeted Village," New York Times, May 31, 1999.
    96 Human Rights Watch interview, Stenkovac II refugee camp, Macedonia, May 9, 1999.
    97 OSCE/ODIHR, Kosovo/Kosova: As Seen, As Told, Part I, pp. 94-96.
    98 Physicians for Human Rights, War Crimes in Kosovo: A Population-Based Assessment of Human Rights Violations Against Kosovar Albanians, August 1999.
    99 International Campaign to Ban Landmines, Landmine Monitor Report 2000, August 2000.
    100 "103 Killed and 394 Disabled by Mines," KosovaLive, November 20, 2000.
    101 Report of the Secretary General on the United Nations Interim Administration Mission in Kosovo, June 6, 2000 (S/2000/538).
    102 Status Report, United Nations Interim Administration Mission in Kosovo, July 14, 2000.
    103 Lord Robertson of Port Ellen, secretary general of NATO, "Kosovo One Year One-Achievement and Challenge," March 21, 2000.
    104 Basic Points of the Statement by the Representative of the Federal Republic of Yugoslavia at the International Seminar on Anti-Personnel Mines, Budapest, March 26-28, 1998. For details of statement and the FRY government's position, see Human Rights Watch, Humanitarian Law Violations in Kosovo.
    105 International Campaign to Ban Landmines, Landmine Monitor Report 2000, August 2000.
    106 Consolidated Minefield Survey Results: Kosovo, the HALO Trust, Pristina, August 14, 1999.
    107 Kosovo Crisis Fact Sheet #133, U.S. Agency for International Development, December 10, 1999.
    108 Military Technical Agreement Between the International Security Force ("KFOR") and the Governments of the Federal Republic of Yugoslavia and the Republic of Serbia, Article II, Paragraph 2.
    109 Press Briefing by the Secretary-General's Special Representative for Humanitarian Affairs in Kosovo, September 29, 1999.
    110 See Human Rights Watch, "Ticking Time Bombs: NATO's Use of Cluster Munitions in Yugoslavia," A Human Rights Watch Report, vol. 11, no. 6 (D), June 1999.
    111 See Human Rights Watch, "Civilian Deaths in the NATO Air Campaign," A Human Rights Watch Report, vol. 12, no. 1 (D), February 2000.
    112 International Campaign to Ban Landmines, Landmine Monitor Report 2000, August 2000. See also Human Rights Watch, "Civilian Deaths in the NATO Air Campaign," Cluster Bombs: Memorandum for CCW Delegates, December 16, 1999, and "Ticking Time Bombs: NATO's Use of Cluster Munitions in Yugoslavia."
    113 Estimates of overall dud rates vary from the conservative 2 to 5 percent claimed by manufacturers, to up to 23 percent observed in acceptance and operational testing, to some 10 to 30 percent observed on the ground in areas of Iraq after the Gulf War. Human Rights Watch has used a conservative estimate of 5 percent mechanical and fuse failures to estimate the humanitarian effect. For details, see Human Rights Watch, "Ticking Time Bombs: NATO's Use of Cluster Munitions in Yugoslavia."
    114 NATO, "Teaching Kids to Stay Away-Cluster Bomb Fact-Sheet,", July 6, 2000.
    115 BBC News, "Kosovo Mine Expert Criticizes NATO," May 23, 2000.