To ascertain whether male circumcision might explain some of the geographical variation in human immunodeficiency virus (HIV) seroprevalence in Africa, we investigated the association between the practice of male circumcision at a societal level and HIV seroprevalence. Male circumcision practices for over 700 African societies were identified, and HIV seroprevalence in general adult populations from 140 distinct locations in 41 countries was obtained. In locations where male circumcision is practised, HIV seroprevalence was considerably lower than in areas where it is not practised. This study supports the hypothesis that lack of circumcision in males is a risk factor for HIV transmission.