Experimental carcinogenesis using fish species as alternative models is a dynamic field of research. The 1940's expansion of synthetic chemical producing industries coincided with a number of pollution-associated fish neoplasia epizootics, with polycyclic aromatic hydrocarbons as significant components of contaminated sediment in several cases. Epizootics of primarily liver and skin neoplasia in benthic species near coastal urban or industrial areas indicated the sensitivity of fish species to known mammalian carcinogens. Stressing a mechanistic approach, investigators have used data compiled from epizootics as the backbone of current research efforts to define carcinogenesis in fish species. With liver as the focus, patterns of neoplastic development similar to those seen in rodent bioassays have been induced in various fish species by genotoxic carcinogens. Similarities between fish and rodent models include chemical and species-specific responses to exposure and the development of predictable preneoplastic and neoplastic lesions. The expression of molecular molecules related to carcinogenesis is currently under investigation, which includes alterations in certain proteins, enzyme activity, and oncogene/tumor suppressor gene function. The potential for the application of research findings to both human and environmental health issues makes fish species attractive and valuable alternative models in carcinogenesis and toxicity research.