OBJECTIVE Most twin and adoption studies of alcoholism have ascertained cases through treatment settings or archival data; these subjects may differ from affected subjects identified epidemiologically. The authors studied the importance of genetic influences on risk of alcohol-related disorders in a new population-based twin sample. METHODS Structured personal interviews were used to assess DSM-III-R-defined and DSM-IV-defined alcohol abuse and dependence among 3,516 twins from male-male pairs born in Virginia between 1940 and 1974. RESULTS The magnitude of resemblance among twin pairs was similar across several definitions of alcoholism and was substantially higher among 861 identical pairs than among 653 fraternal pairs. On the basis of a liability threshold model, 48%-58% of the variation in liability was attributed to additive genetic factors, with the remainder attributed to environmental influences not shared by family members. When a treatment-based proband concordance model was used, evidence for shared environmental as well as genetic influences emerged. CONCLUSIONS In this first population-based study of male twins from the United States, it was found that genetic factors played a major role in the development of alcoholism among males, with similar influence for alcohol abuse and alcohol dependence. Prior findings implicating the influence of common environment may be attributable to sampling strategy; in this population-based sample, environmental factors shared by family members appear to have had little influence on the development of alcoholism in males.