Design and methods: We examined alcohol use before sex and incident HIV in a population-based cohort in Rakai, Uganda, between 1994 and 2002. Adjusted incidence rate ratios (adjIRR) of HIV acquisition and 95% confidence intervals (CI) were estimated by Poisson multivariate regression. We also estimated adjusted prevalence rate ratios to assess the association between alcohol use and the number of sex partners and consistency of condom use.

Results: In 6791 men and 8084 women HIV incidence was 1.4 per 100 person-years and 1.5 per 100 person-years, respectively. After adjustment for sociodemographic and behavioral factors, the risks of HIV when one partner consumed alcohol before sex were: adjIRR 1.67, 95% CI 1.17–2.40 among men, and adjIRR 1.40, 95% CI 1.02–1.92 among women, and when both partners consumed alcohol the risks were adjIRR 1.58, 95% CI 1.13–2.21 among men, and adjIRR 1.81, 95% CI 1.34–2.45 among women. Alcohol use was significantly associated with inconsistent condom use and multiple sexual partners in both sexes.

Conclusion: The use of alcohol before sex increases HIV acquisition. A reduction of alcohol use should be incorporated into HIV prevention programmes.