OBJECTIVES: This paper reviews the potential intervention approaches that have been used to prevent HIV through reducing risky sexual behaviours, summarizes the evidence on their effectiveness, primarily from trials, and discusses the way forward both in terms of research and programmes. METHODS: An update of a recent systematic review of HIV prevention interventions, focusing on trials that have included HIV as an outcome. RESULTS: Five major intervention approaches have been used: community-wide sexual health education, adolescent sexual health interventions, interventions among groups most at risk, promotion of HIV testing and counseling, and interventions among HIV-positive individuals. The latter have often been underemphasized in programmes and research. Effective targeting of interventions to prevent HIV acquisition requires an understanding of HIV incidence by age and sex, whereas HIV prevalence patterns are critical for targeting interventions to reduce HIV transmission (positive prevention). Unfortunately, none of the nine behavioural randomized trials with HIV outcomes have shown a significant impact on HIV. Sometimes this has clearly been due to issues in trial design such as inadequate sample size or low coverage, but not always. Although more encouraging, trials with behavioural outcomes only cannot be used to assume an impact on HIV due to the potential for misreporting and biases in reported sexual behaviour. CONCLUSION: Future research and programmes should place greater emphasis on interventions to reduce HIV transmission as well as acquisition, the sexual norms of the wider population, include a focus on concurrency, and on greatly increasing community acceptance and use of condoms.