The total number, rates, and causes of mortality in Kosovo during the last war remain unclear despite intense international attention. Understanding mortality that results from modern warfare, in which 90% of casualties are civilian, and identifying vulnerable civilian groups, are of critical public-health importance. In September 1999 we conducted a two-stage cluster survey among the Kosovar Albanian population in Kosovo. We collected retrospective mortality data, including cause of death, for the period of the conflict. The survey included 1197 households comprising 8605 people. From February, 1998, through June, 1999, 67 (64%) of 105 deaths in the sample population were attributed to war-related trauma, corresponding to 12,000 (95% CI 5500-18,300) deaths in the total population. The crude mortality rate increased 2.3 times from the pre-conflict level to 0.72 per 1000 a month. Mortality rates peaked in April 1999 at 3.25 per 1000 a month, coinciding with an intensification of the Serbian campaign of "ethnic cleansing". Men of military age (15-49 years) and men 50 years and older had the highest age-specific mortality rates from war-related trauma. However, the latter group were more than three times as likely to die of war-related trauma than were men of military age (relative risk 3.2). Raising awareness among the international humanitarian community of the increased risk of mortality from war-related trauma among men of 50 years and older in some settings is an urgent priority. Establishing evacuation programmes to assist older people to find refuge may prevent loss of life. Such mortality data could be used as evidence that governments and military groups have violated international standards of conduct during warfare.