Fourteen long-term toxicity studies were reviewed in an effort to evaluate the potential carcinogenic activity of styrene and styrene oxide in animals. Each study was reviewed and evaluated for detail and adequacy of design, adequacy of reported data and interpretation. The results of the review are: 1. There is no convincing evidence for a carcinogenic action of styrene in animals, even though it has been studied in several species and by several routes of exposure: inhalation, gavage, in the drinking-water and by intraperitoneal and subcutaneous injection. Most of the studies of styrene, however, have deficiencies in design and/or conduct. 2. Styrene oxide was carcinogenic to the forestomach of rats and mice of each sex after exposure by gavage at all doses tested, including one as low as 50 mg/kg per day. An increase in the incidence of liver neoplasms was observed in male mice in one study. No carcinogenic activity was observed in mice exposed by skin painting. The relevance to humans of the studies in which exposure was by gavage is limited because: (i) the route is less than ideal for extrapolating to human risk from exposure by inhalation or dermally; (ii) xenobiotics often cause neoplasms at this site when given at high concentrations; and (iii) neoplasms at sites distant from the site of exposure were found in only one sex of one species. 3. None of the studies of styrene or styrene oxide reported here is well suited for extrapolating to potential carcinogenic activity in humans, because all have deficiencies in design, conduct and/or interpretation. An up-to-date chronic inhalation study would have to be conducted in order to evaluate this aspect of hazard assessment.