Stereoselective metabolism of mephenytoin has been investigated in four normal subjects by comparing urinary recoveries of hydroxylated metabolites after administration of racemic RS-mephenytoin (1.4 mmol/day) and R-mephenytoin (0.7 mmol/day) on separate occasions. Gas chromatography-mass spectrometry was employed to measure the urinary recovery of 3-methyl-5-(4-hydroxyphenyl)-5-ethylhydantoin (4-OH-M) and mephenytoin catechol, methylcatechol, and dihydrodiol metabolites. Following a single oral dose of racemic mephenytoin, 4-OH-M, mephenytoin catechol, and methylcatechol metabolites were identified in urine mainly as conjugates, whereas the dihydrodiol metabolite was recovered mainly in its unconjugated form. Urinary elimination of each metabolite was similar on days 1 and 10 of chronic racemic mephenytoin administration. Following R-mephenytoin administration, urinary recoveries of hydroxylated metabolites were five to 10 times smaller than after administration of the racemic drug. This implies substrate-stereoselective hydroxylation of the S-enantiomer of mephenytoin. In one subject with a genetic deficiency of aromatic mephenytoin hydroxylation deficiency, the excretion of each hydroxylated mephenytoin metabolite after RS-mephenytoin administration was decreased to 5-15% of the values found in the four extensively hydroxylating study volunteers. The impaired formation of hydroxylated mephenytoin metabolites in genetic hydroxylation deficiency, in conjunction with stereoselective hydroxylation of S-mephenytoin via an extensive NIH shift in normal man, is consistent with the hypothesis that the formation of the S-mephenytoin arene oxide is under genetic control and represents the initial enzymatic reaction of stereoselective aromatic mephenytoin hydroxylation. The formation of this potentially reactive metabolite of S-mephenytoin may have implications in mephenytoin-induced toxicity.