OBJECTIVE: Misreporting of circumcision status may affect observed relationships between circumcision status and HIV or other sexually transmissible infections. As no data exist on the validity of self-reported circumcision status among homosexual men, we investigated the agreement between self-report and examination findings in a subgroup of participants in the Health in Men (HIM) study in Sydney, Australia. METHODS: A subgroup of 240 participants in the community based HIM cohort study attending annual interview agreed to a brief genital examination by a trained study nurse who was unaware of their previous self-reported circumcision status. RESULTS: Five participants reported being uncircumcised at baseline but were classified as circumcised on examination. All participants who self-reported being circumcised were found on examination to be circumcised. Three cases in which the examining study nurse was unsure of participants' circumcision status were excluded. Of the remaining 237 participants, 155 (65.4%) were classified as circumcised on examination, including five men who self identified as uncircumcised. Compared with examination, self-reported circumcision status resulted in a sensitivity of 96.8%, specificity of 100%, positive predictive value of 100% and negative predictive value of 94.3%. The overall agreement between circumcision status on examination and self-report was 97.9% (kappa score, 0.95; p<0.001) CONCLUSION: Self-report was a valid measure of circumcision status in this group of predominantly Anglo gay-community-attached men. We believe our findings can be generalised to similarly aged gay-community-attached men in other developed countries.