Identical and fraternal human twins are a unique resource for studies of the origin and natural history of various diseases, because their genetic similarities are well defined and easily understood. Most investigations using twins fall into two basic categories. In one, genetic factors that contribute to the cause of disease are assessed directly. In the other category, hereditary factors are excluded so that the etiologic importance of nongenetic factors can be evaluated in comparisons that control genetic variation. Both kinds of studies have helped to confirm or dispel specific hypotheses, have provided insights into disease processes, and have shed new light on the usefulness of this study population for future research. Laboratory and clinical investigators assign subjects randomly to different treatment groups to control and evaluate perturbations in their data that are due to factors not specifically accounted for in the study design. In studies on nonhuman subjects, littermates of inbred strains are often randomly assigned to treatment groups to reduce as much as possible the extraneous effects of genetic and early environmental factors. Research on human beings must accommodate a number of constraints: human beings are genetically heterogeneous; they may not be assigned to potentially harmful treatments; and they cannot be assigned at random in studies of spontaneously occurring outcomes, such as diseases of unknown cause. Human twins, however, provide opportunities for evaluating genetic and environmental factors in pairs of subjects of the same age who are either genetically identical or closely related and who have usually shared a common family background and many childhood environmental exposures and experiences.