OBJECTIVE To assess risk factors for hepatitis A infection among homosexual and bisexual men during a community-wide outbreak of hepatitis A in New York City. METHODS Twenty-five homosexual and bisexual men, 20 to 49 years of age with hepatitis A identified from health department surveillance data (cases) were compared with 42 homosexual and bisexual men of similar age distribution who were seronegative for hepatitis A virus and identified from private physician offices (controls). Odds ratio (OR) were determined for acute hepatitis A infection according to demographics, numbers of sexual partners, frequency of specific sexual behaviors, and self-reported human immunodeficiency virus status. RESULTS Cases had more anonymous sex partners (0 to 1 partner versus > 1 partner) than controls during the 6 weeks before illness onset (OR = 4.4, 95% confidence interval [CI] 1.4 to 14.4). Cases were more likely than controls to have engaged in group sex (OR = 3.8, 95% CI 1.1 to 12.6). Among specific sexual behaviors examined, oral-anal intercourse (oral role) and digital-rectal intercourse (digital role) with anonymous sex partners were more commonly reported by cases than controls (OR = 9.7, 95% CI 1.2 to 78.7 and OR = 2.6, 95% CI 1.0 to 7.4, respectively). Multivariate analysis showed that > 1 anonymous sex partner, group sex, oral-anal intercourse, and digital-rectal intercourse were associated with illness in models controlling for duration of sexual activity. Because these variables were highly correlated, independent risk could not be evaluated in a single model. CONCLUSIONS Hepatitis A infection among homosexual and bisexual men is associated with oral-anal and digital-rectal intercourse, as well as with increasing numbers of anonymous sex partners and group sex. These findings reinforce the importance of developing educational activities for homosexual and bisexual men that focus on risk reduction for hepatitis A as well as other sexually transmitted disease spread via the fecal-oral route.