The mechanisms by which polycyclic aromatic hydrocarbons (PAHs) induce mutations and cancer have been the subject of considerable attention for several years. Various theoretical and experimental models have been advanced to explain the effects of structural variations on mutagenicity and carcinogenicity. Among these are the BAY region hypothesis developed by Jerina and coworkers, the K and L region hypotheses first proposed by Pullman and Pullman, the radical cation hypothesis, and others. One of the most widely acceptable today is the BAY region hypothesis, based originally on studies of the carcinogenicity of benzo(a)pyrene, that finds the epoxide formed in the BAY region is more carcinogenic than the parent compound itself. This hypothesis has since been extended to other PAHs, and efforts are underway to identify the specific physicochemical characteristics of the region that make it reactive.