A retrospective study was made of patients in Sheffield who had had repeated gonorrhea infections during 1976-1979. Repeaters--defined as individuals with at least one gonococcal reinfection within 12 months of their index infection in any year--were a constant proportion of the absolute number of individuals infected in any calendar year. Thus, an average of 18.2% of individuals were repeaters who contributed 30.3% of the annual number of heterosexually acquired gonococcal infections. Repeaters of either sex were more likely to be young, unmarried, unemployed, and black; they more commonly had a history of preceding sexually transmitted infections, other sexually transmitted diseases accompanying their gonorrhea, and gonococcal isolates relatively resistant to penicillin. Discriminant analysis of their characteristics suggests that potential repeaters might be predicted. It is hypothesized that the local incidence of gonorrhea is directly proportional to the number of repeaters, and it is felt that control endeavors should be focused on potential repeaters if the incidence of this disease is to be reduced.