It has become widely accepted that metabolic drug-drug interactions can be forecast using in vitro cytochrome P450 (CYP) data. For any CYP form-inhibitor pair, the magnitude of the interaction will depend on the potency of the inhibitor (inhibition constant, Ki) the concentration of the inhibitor available for inhibition ([I]), the fraction of the substrate dose metabolized by CYP (fm), and the fraction of the CYP-dependent metabolism catalyzed by the inhibited CYP form (e.g., fm,CYP3A4). While progress is being made toward our understanding of the factors necessary for predictions of [I]/Ki in vivo, it is evident that there is a need for quantitative databases that contain in vitro (e.g., Ki, fm,CYP3A4) and in vivo pharmacokinetic/absorption-distribution-metabolism-excretion (PK/ADME) data (e.g., fm) for a large number of marketed drugs. Ultimately, such databases would allow one to integrate all of the data necessary for the prediction of drug-drug interactions and permit the rational evaluation of new drug entities.