Consistent with expectations based on human in vitro microsomal experiments, administration of fluconazole (400 mg/day) for 6 days to six human volunteers significantly reduced the cytochrome P450 (P450)-dependent metabolic clearance of the warfarin enantiomers. In particular, P4502C9 catalyzed 6- and 7-hydroxylation of (S)-warfarin, the pathway primarily responsible for termination of warfarin's anticoagulant effect, was inhibited by approximately 70%. The change in (S)-warfarin pharmacokinetics caused by fluconazole dramatically increased the magnitude and duration of warfarin's hypoprothrombinemic effect. These observations indicate that co-administration of fluconazole and warfarin will result in a clinically significant metabolically based interaction The major P450-dependent, in vivo pathways of (R)-warfarin clearance were also strongly inhibited by fluconazole. 10-Hydroxylation, a metabolic pathway catalyzed exclusively by P4503A4, was inhibited by 45% whereas 6-, 7-, and 8-hydroxylations were inhibited by 61, 73, and 88%, respectively. The potent inhibition of the phenolic metabolites suggests that enzymes other than P4501A2 (weakly inhibited by fluconazole in vitro) are primarily responsible for the formation of these metabolites in vivo as predicted from in vitro kinetic studies. These data suggest that fluconazole can be expected to interact with any drug whose clearance is dominated by P450s 2C9, 3A4, and other as yet undefined isoforms. Overall, the results strongly support the hypothesis that metabolically based in vivo drug interactions may be predicted from human in vitro microsomal data.