The use of in vitro systems in the assessment of xenobiotic metabolism has distinct advantages and disadvantages. While isolated hepatocytes and microsomes prepared from human liver may be used to generate data for comparisons among species and in vitro systems, such comparisons are generally performed on the basis of microsomal protein or million (viable) hepatocytes. Recently, in vitro data have been investigated for their value as quantitative predictors of in vivo metabolic capacity. Because of the existence of large amounts of trichloroethylene (TRI) data in the human, we have examined the metabolism of TRI as a case study in the development of a method to compare metabolism across species using in vitro systems and for extrapolation of metabolic rates from in vitro to in vivo. TRI is well metabolized by human hepatocytes in culture with a K(m) of 266 +/- 202 ppm (mean +/- SD) in headspace and a Vmax of 16.1 +/- 12.9 nmol/h/10(6) viable hepatocytes. We determined that human liver contains approximately 116 x 10(6) hepatocytes and 20.8 mg microsomal protein/g, based on DNA recovery and glucose-6-phosphatase activity, respectively. Thus, the microsomal protein content of hepatocytes is 179 micrograms microsomal protein/10(6) isolated hepatocytes. The microsomal apparent Vmax value of 1589 pmol/min/mg microsomal protein extrapolates to 17.07 nmol/h/10(6) hepatocytes. The combination of protein recovery and metabolic rate predicted a Vmax of approximately 1400 nmol/h/g human liver, which, when extrapolated and incorporated into an existing physiologically based pharmacokinetic (PBPK) model for TRI, slightly underpredicted TRI metabolism in the intact human. The quantitation, extrapolation, and inclusion of extrahepatic and cytochrome P450 (CYP)-independent TRI metabolism may increase the predictive value of this approach.