The contribution of hypertension to adult mortality in Africa has not been well studied. Although cross-sectional surveys have provided data on the prevalence of this condition, the relative risk of death associated with hypertension has not been defined. In the face of high levels of competing mortality from infectious disease among the general population, and the virtual absence of atherosclerotic precursors, estimates of risk derived from industrialised countries may not be generalisable to this setting. We conducted a 2-year prospective study among 1344 mean and women in a rural community in south-western Nigeria. The prevalence of hypertension (140/90 mm Hg) at baseline was 9.3%. In the observational phase, 3.0% of the survey participants died each year. Among the 74 decedents, hypertension was nearly twice as common as among those who survived (14.9% vs 8.4%). In multivariate analysis the risk of death increased over 60% for a 20 mm Hg increase in diastolic blood pressure. The population attributable risk, or the reduction in mortality that would have been observed if hypertension were not present in this community, was estimated as 7%. These findings document an identifiable impact of hypertension on all-cause mortality in rural Africa and demonstrate that programs to evaluate potential treatment options are needed.