This review describes statistical models for the biological interaction of susceptibility genes and environmental exposures, as observed in epidemiologic studies. The importance of metabolic transformation of industrial carcinogens and the potential role of genetic polymorphisms in metabolic enzyme activity are outlined. Several genetic polymorphisms have been associated with cancer risk, but the link with the relevant exposures has only infrequently been specified. For example, studies show that slow N-acetylation increased bladder cancer risk among workers exposed to some arylamines, as found among dye workers in England, but that this effect does not hold for benzidine exposure. This link of a genetic susceptibility factor with cancer risk due to some aromatic amines, but not to others, illustrated the specific nature of metabolic environment/gene interactions. Epidemiologic studies to investigate the role of genetic susceptibility in cancer development promise to further the identification of human carcinogens by focusing on susceptible individuals and, in turn, to enhance understanding of human cancer by relating cancer risk in populations to underlying biologic processes. Occupational studies are key to this effort.