In lifetime bioassays, trichloroethylene (TCE, CAS No. 79-01-6) causes liver tumors in mice following gavage, liver and lung tumors in mice following inhalation, and kidney tumors in rats following gavage or inhalation. Recently developed pharmacokinetic models provide estimates of internal, target-organ doses of the TCE metabolites thought responsible for these tumor responses. Dose-response analyses following recently proposed methods for carcinogen risk assessment from the U.S. Environmental Protection Agency (U.S. EPA) are conducted on the animal tumor data using the pharmacokinetic dosimeters to derive a series of alternative projections of the potential carcinogenic potency of TCE in humans exposed to low environmental concentrations. Although mechanistic considerations suggest action of possibly nonlinear processes, dose-response shapes in the observable range of tumor incidence evince little sign of such patterns. Results depend on which of several alternative pharmacokinetic analyses are used to define target-organ doses. Human potency projections under the U.S. EPA linear method based on mouse liver tumors and internal dosimetry equal or somewhat exceed calculations based on administered dose, and projections based on mouse liver tumors exceed those from mouse lung or rat kidney tumors. Estimates of the carcinogenic potency of the two primary oxidative metabolites of TCE--trichloroacetic acid and dichloroacetic acid, which are mouse liver carcinogens in their own right--are also made, but it is not clear whether the carcinogenic potency of TCE can be quantitatively ascribed to metabolic generation of these metabolites.