The aim of this study was to explore the potential of recombinant cytochrome P450 (P450) enzymes for human metabolic clearance prediction. The relative abundance and relative activity approaches were compared as methods to bridge the gap between catalytic activities in recombinant P450 enzymes and human liver microsomes (HLMs). Relative activity factors were measured by determining the intrinsic clearance (CL ) of probe substrates (bu- furalol-CYP2D6, diclofenac-CYP2C9, midazolam-CYP3A4, and phenacetin-CYP1A2) in recombinant P450s and 16 HLM donors. Simultaneous determination of drug depletion and metabolite formation profiles has enabled a direct comparison of these methods for CL determination. Of the 110 drugs tested, 66% were metabolized by one or more P450 enzymes; of these 44% of were metabolized by CYP3A4 (0.3-21 fil/min/pmol of P450), 41% by CYP2D6 (0.6-60 ft l/min/pmol of P450), 26% by CYP2C19 (0.4-8.1 fil/min/pmol of P450), 9% by CYP1A2 (0.4-2.5 fil/min/pmol of P450), and 4% by CYP2C9 (0.9-6.4 ft l/min/pmol of P450). Recombinant enzymes demonstrated improved prediction reliability relative to HLMs and hepatocytes. The most reliable correlations in terms of lowest bias and highest precision were observed by comparing in vivo CL , calculated using the parallel-tube model and incorporating fraction unbound in blood, with in vitro CL determined using relative activity factors and adjusted for nonspecific binding. Predictions were less reliable using the relative abundance approach. For these drugs, recombinant P450 enzymes offer improved assay sensitivity compared with HLMs and cryo- preserved hepatocytes for CL determination using the drug depletion method.