Background:  Prior studies suggest that male circumcision reduces the risk of HIV acquisition in men.

Methods:  We conducted a randomized trial in rural Rakai District, Uganda, and enrolled 4996 uncircumcised HIV^– men, aged 15 to 49 years who agreed to receive their HIV results and counseling. Men were randomized to immediate circumcision (2474), or circumcision delayed for 24 months (2522), and were followed-up at 6, 12, and 24 months. HIV incidence rate ratios (IRR) and 95% confidence intervals (95%CI) of HIV were estimated by Poisson multivariate regression, and prevalent rate ratios (PRR) of sexually transmitted infection symptoms were estimated by log-binomial regression. The trial was stopped early, after accrual of 72% the projected total information. We also assessed circumcision safety in HIV^– men enrolled in this trial, and in HIV⁺ men enrolled in a parallel trial.

Results:  Intervention and control men had comparable characteristics at enrollment.  Retention rates were between 88 and 90% in both arms. In the intent-to-treat analysis, the cumulative HIV incidence over 24 months was 0.66/100 person-years (22/3352.4 person-years) in the intervention arm and 1.33/100 person-years (45/33,918 person-years) in the control arm (IRR 0.49, 95%CI 0.30 to 0.82; p = 0.007). The as treated IRR was 0.45, 95%CI 0.25 to 0.78; p = 0.002). Among sexually active men, HIV incidence was lower in the intervention arm for all covariates. There were no consistent or substantial differentials in sexual risk behaviors between arms. Moderate or severe adverse events related to surgery occurred in 3.6% of circumcised HIV^– men and 3.1% of HIV⁺. All were resolved with treatment.

Conclusions:  Male circumcision reduced HIV incidence and genital ulceration, and there was no consistent evidence of behavioral disinhibition. Circumcision can be promoted for HIV prevention in men. Surgical complications are comparable in HIV^– and HIV⁺ men.