BACKGROUND: Randomized clinical trials have shown that medical male circumcision substantially reduces the risk of contracting HIV. However, relatively little is known about the relationship between traditional male circumcision and HIV risk. This article examines variations in traditional circumcision practices and their relationship to HIV status. METHODS: We used data from the fifth wave of the Cape Area Panel Study (n = 473) of young adults in Cape Town, South Africa, to determine attitudes towards circumcision, whether men were circumcised, at what age, and whether their foreskin had been fully or partially removed. Probit models were estimated to determine the association between extent and age of circumcision and HIV status. RESULTS: There was strong support for traditional male circumcision. 92.5% of the men reported being circumcised, with 10.5% partially circumcised. Partially circumcised men had a 7% point greater risk of being HIV positive than fully circumcised men (P < 0.05) and equal risk compared with uncircumcised men. Most (91%) men were circumcised between the ages of 17 and 22 years (mean 19.2 years), and HIV risk increased with age of circumcision (P < 0.10). CONCLUSIONS: Efforts should be made to encourage earlier circumcisions and to work with traditional surgeons to reduce the number of partial circumcisions. Data on the extent and age of circumcision are necessary for meaningful conclusions to be drawn from survey data about the relationship between circumcision and HIV status.