Previous mouse liver studies with diazepam (DZ), N-desmethyldiazepam (NZ), and temazepam (TZ) confirmed that under first-order conditions, DZ formed NZ and TZ in parallel. Oxazepam (OZ) was generated via NZ and not TZ despite that preformed NZ and TZ were both capable of forming OZ. In the present studies, the concentration-dependent sequential metabolism of DZ was studied in perfused mouse livers and microsomes, with the aim of distinguishing the relative importance of NZ and TZ as precursors of OZ. In microsomal studies, the Kms and Vmaxs, corrected for binding to microsomal proteins, were 34 microM and 3.6 nmole/min per mg and 239 microM and 18 nmole/min per mg, respectively, for N-demethylation and C3-hydroxylation of DZ. The Kms and Vmaxs for N-demethylation and C3-hydroxylation of TZ and NZ, respectively, to form OZ, were 58 microM and 2.5 nmole/min per mg and 311 microM and 2 nmole/min per mg, respectively. The constants suggest that at low DZ concentrations, NZ formation predominates and is a major source of OZ, whereas at higher DZ concentrations, TZ is the important source of OZ. In livers perfused with DZ at input concentrations of 13 to 35 microM, the extraction ratio of DZ (E[DZ]) decreased from 0.83 to 0.60. NZ was the major metabolite formed although its appearance was less than proportionate with increasing DZ input concentration. By contrast, the formation of TZ increased disproportionately with increasing DZ concentration, whereas that for OZ decreased and paralleled the behavior of NZ. Computer simulations based on a tubular flow model and the in vitro enzymatic parameters provided a poor in vitro-organ correlation. The E[DZ], appearance rates of the metabolites, and the extraction ratio of formed NZ (E[NZ, DZ]) were poorly predicted; TZ was incorrectly identified as the major precursor of OZ. Simulations with optimized parameters improved the correlations and identified NZ as the major contributor of OZ. Saturation of DZ N-demethylation at higher DZ concentrations increased the role of TZ in the formation of OZ. The poor aqueous solubility (limiting the concentration range of substrates used in vitro), avid tissue binding and the coupling of enzymatic reactions in liver, favoring sequential metabolism, are possible explanations for the poor in vitro-organ correlation. This work emphasizes the complexity of the hepatic intracellular milieu for drug metabolism and the need for additional modeling efforts to adequately describe metabolite kinetics.