INTRODUCTION: The relevance of circumcision in preventing male-to-male sexual transmission of HIV is poorly understood, in particular because any potential beneficial effect could be diminished by the impact of circumcision on sexual behavior. AIM: We examined the impact of circumcision on sexual experience. METHODS: Univariate and multivariate logistic regressions were performed on data from 1,426 HIV-negative homosexually active men. MAIN OUTCOME MEASURES: We compared the sexual behaviors and preferences of circumcised with uncircumcised men, and men who were circumcised at infancy with those who were circumcised after infancy. RESULTS: Overall, 66% of men (N = 939) in the cohort were circumcised. After adjusting for age and ethnicity, we found no differences between circumcised and uncircumcised men in any insertive or receptive anal intercourse, difficulty using condoms, or sexual difficulties (e.g., loss of libido). Among the circumcised men, we compared those circumcised at infancy (N = 854) with those circumcised after infancy (N = 81). The majority cited phimosis (i.e., an inability to fully retract the foreskin) and parents' decision as the main reasons for circumcision after infancy. After adjusting for age and ethnicity, the men circumcised after infancy were more likely to practice any receptive anal sex (88% vs. 75%, P < 0.05) and to experience erection difficulties (52% vs. 47%, P < 0.05), but less likely to practice any insertive anal sex (79% vs. 87%, P < 0.05) and to experience premature ejaculation (15% vs. 23%, P < 0.05) than those circumcised at infancy. CONCLUSIONS: Our data suggest that overall circumcision status does not affect the HIV-negative gay men's anal sexual behaviors, experience of condom use, or likelihood of sexual difficulties. However, there is some suggestion of differences in sexual practices and preferences among circumcised gay men depending on the age at circumcision. In particular, gay men circumcised later are more likely to engage in and prefer receptive anal intercourse.