Epidemiologic treatment of gonorrhea refers to the administration of antibiotics when the diagnosis is considered likely, but before the results of confirmatory tests are available. Unfortunately, the risk of infection is seldom known. To place epidemiologic treatment on a more rational basis, infection rates were determined for groups of clinic patients defined by easily collected indexes of risk such as reason for attending the clinic, sex, race, and sexual preference, history of contact with gonorrhea, and history of a urethral or vaginal dischange. Infection rates ranged from 0.8% for men seeking marriage licenses to 65.1% for female contacts of men with "established" gonorrhea. By selectively employing gram-stained smears, it is possible to reduce further the need for epidemiologic treatment within the various defined groups of patients. Epidemiologic treatment policies should be determined by each large clinic and should be based on known infection rates for clearly defined groups of patients. Such rates are necessary for obtaining informed consent from patients and for evaluation of the cost-effectiveness of epidemiologic treatment in the overall effort to control gonorrhea.