CYP2C9 mediates the oxidative metabolism of approximately 10% of drugs, some of which are characterized by a narrow therapeutic index. We aimed to validate genotype method and phenotype methodology, for evaluation of CYP2C9 activity in vivo. Thirty-one healthy subjects (22 male) received a single 300 mg dose of phenytoin. Blood was drawn periodically and urine was collected at intervals for 96 h. Plasma phenytoin and 5-(4-hydroxyphenyl)-5-phenylhydantoin (p-HPPH) and urine S and R enantiomers of p-HPPH were determined by high-performance liquid chromatography. CYP2C9 genotyping was obtained by polymerase chain reaction followed by digestion with S au 96I and S ty I for the identification of CYP2C9*2 and CYP2C9*3, respectively. Eighteen subjects were CYP2C9*1 homozygous, seven were CYP2C9*2 heterozygous, four were CYP2C9*3 heterozygous, one was CYP2C9*2 homozygous and one was compound CYP2C9*2/ CYP2C9*3 heterozygous. The allele frequencies of CYP2C9*1, CYP2C9*2 and CYP2C9*3 were 0.76 [95% confidence interval (CI) 0.73–0.79], 0.16 (95% CI 0.13–0.19) and 0.08 (95% CI 0.05–0.11), respectively. The CYP2C9-mediated production of (S)-p-HPPH represented the major metabolic pathway of phenytoin biotransformation as its excretion accounted for 95.6 + 0.9% of ‘total’p-HPPH excretion over the 96 h collection interval. Phenytoin metabolic clearance to produce (S)-p-HPPH (PMC), correlated significantly with (S)-p-HPPH (or ‘total’p-HPPH) content in 0–8, 0–12 and 0–24 urine collections (r = 0.88, 0.85 and 0.89, respectively) and with phenytoin metabolic ratio (PMR) defined as the ratio of urine (S)-p-HPPH (or ‘total’ p-HPPH) to mid-interval plasma phenytoin (r = 0.90, 0.88 and 0.94, respectively). PMC and PMR exhibited a gene–dose effect so that the highest and lowest values were noted in homozygous subjects CYP2C9*1 and subjects carrying two defective alleles, respectively, whereas heterozygous subjects had intermediate values. CYP2C9 genotyping and several phenytoin metabolic indices are correlated with CYP2C9 activity in vivo. The utility of phenytoin to predict the metabolism of other CYP2C9 substrates justifies further evaluation.