Whether the observed association between male circumcision and HIV infection is causal or not has not been verified. We did a meta-analysis of published data and applied Hill's criteria for causality on all available evidence to assess presence of a causal association. Analysis was by the random effects method. Summary estimates were calculated for all studies combined and for sub groups stratified by type of study population, study design, and method of ascertaining circumcision status. Thirteen studies were included. Circumcised men had a reduced risk for HIV infection (adjusted RRoverall = 0.42, 95% CI 0.33-0.53; RR(RCT) = 0.43 95% CI 0.32-0.59, RRobservational = 0.39, 95% CI 0.27-0.56). Available evidence satisfies six of Hill's criteria: strength of association, consistency, temporality, coherence, biological plausibility, and experiment. These results provide unequivocal evidence that circumcision plays a causal role in reducing the risk of HIV infection among men.