The kinetics of oxidation of phenytoin and tolbutamide were determined in freshly isolated hepatocytes and hepatic microsomes from male Sprague-Dawley rats. Similar enzyme kinetic models are applicable to the data from both in vitro systems; a two-site model for phenytoin with a high affinity (Km = 1-5 microM, based on unbound drug concentration), low capacity site and a low affinity, high capacity site, and a one-site model for tolbutamide. Steady-state infusion studies were performed to characterize the Michaelis-Menten parameters for phenytoin disposition in vivo, these data could also be described by a two-site metabolism model (Km 1.3 microM, intrinsic clearance 62 ml/min for unbound drug for the high affinity site). Comparison of in vivo and in vitro parameters (after scaling the latter parameters for either hepatocyte yield or microsomal recovery) showed excellent prediction of in vivo clearance of unbound drug from hepatocyte data (55 ml/min) but underprediction from microsomal data (17 ml/min). In contrast to phenytoin, the in vivo clearance of tolbutamide (1.5 ml/min for unbound drug) was equally well predicted by both hepatocyte (2.4 ml/min) and microsomal (3.1 ml/min) studies. The difference between the utility of in vitro systems to predict the in vivo clearance of these two drugs, which show similar pharmacrokinetic properties (low clearance restricted to unbound drug concentration in blood), may be a consequence of the particular terminal metabolite formed in each in vitro system.(ABSTRACT TRUNCATED AT 250 WORDS)