Lifetime cancer or unit risk estimates for TRI have been calculated by the EPA on the basis of metabolized dose-tumor incidence relationships. Previously, it was common practice to directly extrapolate exposure dose-tumor incidence data from laboratory animal studies to predict cancer risks in humans. Such direct species-to-species extrapolations, however, do not take into account potentially important species differences in systemic uptake, tissue distribution, metabolism, deposition at the site(s) of action, and elimination. The consideration and use of pharmacokinetic and metabolic data can significantly reduce, though not eliminate, uncertainties inherent in species-to-species, route-to-route, and high- to low-dose extrapolations. The total amount of TRI metabolized was considered in the most recent EPA Health Assessment Document for Trichloroethylene to be the effective dose (EFD) producing tumors. Exposure dose-metabolism relationships were determined from direct measurement data in inhalation and oral dosing studies in mice and rats. The magnitude of TRI metabolism in these two species closely approximated body surface area. Thus, it was assumed that the amount of TRI metabolized per square meter of surface area was equivalent among species when calculating human equivalent doses from the animal data. Direct measurement data from an inhalation study in humans were used to calculate the amount of TRI metabolized and the unit risk estimate when a person inhales 1 microgram TRI per cubic meter continuously for 24 h. The EPA Cancer Assessment Group (CAG) elected to use this risk estimate for TRI in air, since it was calculated on the basis of a human metabolized dose rather than unit risk estimates based on animal studies. The current survey of literature and ongoing research uncovered no new animal or human studies in which TRI metabolites were directly measured, which would be any more suitable for use in estimating the total metabolized dose of TRI. On the basis of information now available, it is appropriate to continue to use the total amount of TRI metabolized as the EFD producing tumors in the liver. Use of the total amount metabolized represents an important "step in the right direction" in reducing uncertainties in interspecies extrapolations of data on a chemical such as TRI. TRI is believed to be metabolically activated to a reactive intermediate(s), although the identity of the intermediate(s) is unclear. There is evidence that formation of reactive intermediate(s) and TRI hepatotoxicity are directly proportional to the overall extent of TRI metabolism.(ABSTRACT TRUNCATED AT 400 WORDS)