We compared omeprazole and mephenytoin as probes for the CYP2C19 metabolic polymorphism. Single oral doses of omeprazole (20 mg) or mephenytoin (100 mg) were administered at least 1 week apart to 167 healthy volunteers. Mephenytoin metabolism was measured using the amount of 4'-hydroxymephenytoin and the S/R ratio of mephenytoin in an 8-hour urine collection. Omeprazole hydroxylation was measured using the ratio of omeprazole to 5'-hydroxyomeprazole in serum 2 hours after dosing. All three methods separated poor- or extensive-metabolizer phenotypes with complete concordance. Omeprazole hydroxylation correlated with the S/R ratio of mephenytoin in extensive metabolizers (r2 = 0.681; p < 0.001). Genotyping tests showed that six poor metabolizers of omeprazole were homozygous for a single base pair mutation in exon 5 of CYP2C19. These results support the hypothesis that omeprazole 5'-hydroxylation cosegregates with the CYP2C19 metabolic polymorphism.