Phenytoin (PHT) is primarily metabolized to 5-(4'-hydroxyphenyl)-5-phenylhydantoin (p-HPPH), accounting for 67-88% of an administered dose in humans. p-HPPH is formed by the cytochrome (CYP) 450 enzymes CYP2C9 and CYP2C19, then glucuronidated and excreted into the urine. CYP2C9 catalyses the prochiral formation of (R) and (S)-p-HPPH, and is approximately 40 times more stereoselective towards the formation of the (S) isomer whereas CYP2C19 is not stereoselective. Because of differential stereoselectivity, polymorphisms in the genes can alter the (S)/(R)-p-HPPH ratios. Genotyping for CYP2C9 and CYP2C19 was accomplished by a Taqman based assay. Twelve and twenty-four hour urine samples were collected from 45 epilepsy patients taking PHT under steady-state conditions and (S)/(R) ratios of p-HPPH were determined by chiral HPLC separation. The mean urinary (S)/(R) ratio in the 12-24h urine collection in subjects homozygous for CYP2C9*1/*1, CYP2C19*1/*1 was 24.2+/-3.1(n=21), whereas ratios in CYP2C9*1/*2 and CYP2C9*1/*3 subjects, were 11.1+/-3.3(n=7) and 2.7+/-0.6(n=2), respectively. One CYP2C9*2/*3 patient had a ratio of 2.1. Unexpectedly, CYP2C9*1/*1, CYP2C19*1/*2 subjects had a mean (S)/(R) ratio as low as 12.9+/-1.7(n=12). Our results are generally consistent with single dose PHT studies. However, the (S)/(R)-p-HPPH ratios for the CYP2C9*1/*1, CYP2C19*1/*2 subjects, expected to be in the range of 30-40, were only 12.9, suggesting some undetected linkage disequilibrium between CYP2C9 and CYP2C19 genes that could affect PHT elimination. Furthermore, our study suggests that measurement of urine ratios cannot be used as a marker for genotype determination.