Epidemiologic methods are designed to identify risk factors involved in production of a particular disease, even when the specific etiologic agent is unknown. However, a major problem currently presenting obstacles to research on the acquired immunodeficiency syndrome (AIDS) relates to classification; approaches are described that will permit reclassification when better laboratory techniques are developed or when more information is available from studies on natural history. Further descriptive studies will be valuable in the determination of whether extension of disease has occurred to new population groups or geographic areas. Case control studies can provide information on changes of known risk factors or can confirm and help to specify those factors already recognized. Because of the dynamic nature of AIDS and the probable existence of multiple risk factors, prospective cohort studies will be of greatest value, although they will take longer to complete. As understanding of the etiologic factors improves, preventive measures will become more efficient.