OBJECTIVE: To investigate the relationship between nonoccupational postexposure prophylaxis (NPEP) use and future HIV risk behaviours and HIV infection in a cohort of HIV-negative homosexual men in Sydney, Australia. DESIGN: Prospective analysis of NPEP use in a community cohort from 2001 to 2007. METHODS: In the Health in Men study cohort, men were annually questioned about NPEP use and tested for HIV. Every 6 months, detailed quantitative data on unprotected anal intercourse were collected. Cox regression models examined risk factors for incident NPEP use, HIV seroconversion and time trends in NPEP use. The change in the number of unprotected anal intercourse acts with nonseroconcordant partners before and after NPEP was examined using the Wilcoxon signed-rank test. RESULTS: One thousand four hundred and twenty-seven participants were enrolled. At baseline, 78.5% of participants had heard of NPEP, which increased to 97.4% by the fifth annual interview. NPEP use increased significantly from 2.9 per 100 person-years in 2002 to 7.1 per 100 person-years in 2007 (P = 0.007). Unprotected anal intercourse was a strong predictor of incident NPEP use. Use of NPEP was not associated with changes in HIV risk behaviour. Men who received NPEP had a significantly higher rate of subsequent HIV seroconversion (hazard ratio 2.67, 95% confidence interval 1.40-5.08, P = 0.003). CONCLUSION: Awareness of the availability of NPEP in this cohort was nearly universal. Use was common and increased rapidly over the study period. NPEP was targeted mostly towards high-risk behaviours. Use of NPEP was not associated with reductions in risk behaviour, and men who received NPEP continued to be at high risk of subsequent HIV infection.