In a study of 689 male homosexuals 290 (42%) were found to have antibodies to hepatitis A virus. The 399 men who did not have antibodies were followed up for up to 690 days, and 35 cases of hepatitis A were detected. The attack rate at the end of the study was 14%. The incidence climbed steadily, indicating that the cases of hepatitis A did not occur in clusters. Statistical analysis showed that the prevalence of antibodies to hepatitis A virus was significantly correlated with the duration of homosexual activity (p less than 0.006), and this was independent of age. The incidence of hepatitis A was found to be correlated with the number of different sexual partners in the preceding six months. It is concluded that hepatitis A is a sexually transmitted disease among homosexual men in countries with a low rate of exposure to hepatitis A during childhood.