We examined the association between education and prevalent HIV-1 infection in the Rakai district, rural Uganda based on a cross-sectional analysis of a population-based cohort. In 1990, 1397 men and 1705 women aged 13 years and older, were enrolled in 31 randomly selected communities. Strata were comprised of main road trading centres, secondary road trading villages and rural villages. Sociodemographic and behavioural data were obtained by interview and serum for HIV serostatus were obtained in the home. The analysis examines the association between sex-specific prevalent HIV infection and educational attainment, categorized as secondary, primary or none. The odds ratios (ORs) and 95% confidence intervals (95% CIs) of HIV infection were estimated, using no education as the referent group. Higher levels of education were associated with a higher HIV seroprevalence in bivariate analyses (OR 2.7 for primary and 4.1 for secondary education, relative to no education). The strength of the association was diminished but remained statistically significant after multivariate adjustment for sociodemographic and behavioural variables (adjusted OR of HIV infection 1.6 (95% CI: 1.2-2.1)) for primary education and 1.5 (95% CI: 1.0-2.2) for secondary education. Stratified multivariate analyses by place of residence indicated that the association between education and HIV prevalence was statistically significant in the rural villages, but not in the main road trading centres and intermediate trading villages. Educational attainment is a significant predictor of HIV risk in rural Uganda, in part because of risk behaviours and other characteristics among better educated individuals. Preventive interventions need to focus on better educated adults and on school-aged populations.