Intrinsic clearance (CL(int)) of seven probe cytochrome P450 substrates, across a wide range of clearance, was compared in microsomes and cryopreserved hepatocytes from the same four livers. Previous comparisons have shown system dependence, but using preparations from different donor livers. Four-fold average underprediction of microsomal CL(int) by hepatocytes (scaled to whole liver) for high clearance substrates (midazolam, nifedipine, and diclofenac) was observed with relatively unbiased prediction (within 1.5-fold average) for the low/medium clearance substrates (tolbutamide, alprazolam, bufuralol, and triazolam). CL(int) of midazolam and nifedipine corresponded between livers over a tenfold range, but the absolute ranges were lower for hepatocytes, indicating independence of hepatocyte bias from substrate. In contrast, the absolute ranges of CL(int) for the low clearance CYP3A4 substrate alprazolam were similar between the systems, indicating independence of hepatocyte bias from enzyme. The trend in CL(int) between the systems was similar to that in a dataset of published CL(int) for 46 substrates in microsomes and hepatocytes (unrelated liver sources), supporting a fundamental rate limitation of the hepatocyte system. A tendency of decreasing V(max) in hepatocytes relative to microsomes, with increasing clearance, suggests that a capacity limitation, such as cofactor rate limitation, may be involved in this phenomenon.