In most countries, pertussis surveillance is inadequate for accurately estimating numbers of cases or deaths. Good estimates are needed to help set priorities for vaccination programmes. We aimed to develop a simple, reliable, and explicit method for estimating pertussis cases and deaths for children under 15 years to calculate the global disease burden in 1999. We estimated the proportion of susceptible children becoming infected in countries with poor vaccination coverage (<70%) in 1999 at 30% by 1 year, 80% by 5 years, and 100% by 15 years of age and for countries with good coverage (> or =70%) at 10% by 1 year, 60% by 5 years, and 100% by 15 years. Vaccine efficacy was estimated at 80% for preventing infection and 95% for preventing deaths. We used UN population estimates and vaccination coverage reported to WHO (adjusted for specific survey data if available). Case fatality ratios for countries with high and low child mortality were derived from published and unpublished work. For some countries with good vital events registration we used reported deaths adjusted for underascertainment. In 1999 there were an estimated 48.5 million pertussis cases in children worldwide. Deaths from pertussis were estimated at 390000 and at 295000 after adjustment for local data sources. Based on this approach, disability-adjusted life years from pertussis (12.7 million) in 2000 exceeded those of other preventable diseases such as lung cancer (11.4 million) and meningitis (5.8 million). This simple approach yields estimates that can be used for setting vaccination programme priorities. Better data are needed on the public health importance of pertussis in high mortality countries, the benefits of incomplete vaccination, and the harm from delayed vaccination.