We analyzed data from a multisite study of 1,063 gay or bisexual men attending sexually transmitted disease clinics to evaluate factors predicting failure to disclose human immunodeficiency virus (HIV) risk behaviors to clinic staff and the extent of such failure. We compared data from a brief screening assessment on unprotected anal and oral sex with data on the same behaviors from a subsequent detailed interview. We also compared behavioral data from screening and the interview with data on diagnoses of rectal gonorrhea abstracted from medical charts. Of 523 men reporting unprotected anal sex at interview, 29% failed to report this behavior at screening. Men failing to disclose unprotected anal sex were also less likely to disclose engaging in unprotected oral sex. Among men reporting no unprotected anal sex, either at screening or interview, 1.6% were diagnosed with rectal gonorrhea. Logistic regression analyses comparing men who did and did not disclose at screening having engaged in unprotected anal sex showed that men who failed to disclose reported greater involvement in gay organizations, greater perceived peer support for condoms, fewer episodes of unprotected anal sex in the last four months, and lower rates of substance abuse treatment. Our data suggest that men who failed to disclose may have lower risk levels, and may be more integrated into the gay community. Brief interviews, as opposed to detailed ones, also may underestimate incidence of unsafe sex. Where feasible, HIV risk assessment and counseling and laboratory screening should be routinely provided to all clinic attendees, regardless of self-reports.