Lifetime exposure to perchloroethylene by inhalation has been shown to cause a low incidence of renal tumors in male rats. The mechanisms responsible for the induction of these tumors have been investigated following exposure of rats to perchloroethylene by oral gavage (1500 mg/kg for up to 42 days) or by inhalation (400 ppm for 28 days). Comparisons have been made between rats and mice in vivo and between rats, mice, and humans in vitro. High doses of perchloroethylene given by gavage have been shown to be toxic to the rat kidney, causing increases in urinary markers of kidney damage. A marked accumulation of protein droplets (alpha-2u-globulin) was seen in the P2 segment of the kidney proximal tubules. This response were not seen after inhalation exposure to 400 ppm perchloroethylene for 28 days and hence may not be associated with the tumors seen at this dose level. Protein droplet formation was seen after exposure to 1000 ppm perchloroethylene, suggesting that 400 ppm is below the threshold dose required to induce this response. Perchloroethylene has been shown to be metabolized by glutathione conjugation in the liver, resulting in the formation of a mutagenic cysteine conjugate which is activated by the kidney enzyme beta-lyase. Levels of the mercapturic acid of perchloroethylene have been compared in rat and mouse urine. The enzyme kinetics of hepatic glutathione conjugation and renal beta-lyase activation have been compared in rat, mouse, and human tissues in vitro. Results of these studies are consistent with the rat being the species susceptible to kidney tumors. Although human kidney was shown to contain beta-lyase, glutathione conjugation of perchloroethylene could not be detected in human liver. Perchloroethylene-induced male rat kidney tumors may be a result of chronic toxicity, protein droplet nephropathy, and genotoxicity from the beta-lyase pathway. These mechanisms appear to have little relevance to humans.