A single oral dose of omeprazole (20 mg) was given orally to 160 healthy Caucasian Swedish subjects and tested as a probe for CYP2C19. The study was nonrandomized and included seven subjects previously classified as poor metabolizers (PM) of S-mephenytoin. The ratio between the plasma concentrations of omeprazole and hydroxyomeprazole (metabolic ratio; MR) was determined by HPLC in a blood sample drawn 3 h after drug intake. In 17 subjects the test was repeated and the MRs of omeprazole on the two occasions were correlated (rs = 0.85; p < 0.0001). There was a significant correlation between the MR of omeprazole and the S/R mephenytoin ratio among 141 subjects, in whom both ratios were determined (rs = 0.63, p < 0.001). All seven PMs of S-mephenytoin had higher MRs of omeprazole (7.1-23.8) than extensive metabolizers (EM) (0.1-4.9). All 160 subjects and another 15 Caucasian Swedish PMs previously phenotyped with mephenytoin were analysed with respect to the presence of the CYP2C19m1 allele by PCR amplification of the intron 4/exon 5 junction followed by Sma I digestion. EMs heterozygous for the CYP2C19m1 gene had MRs of omeprazole and S/R ratios of mephenytoin that were higher than those of subjects who were homozygous for the wild-type allele (p = 0.0001). Nineteen of the 22 PMs were homozygous for the CYP2C19m1 gene. Three were heterozygous for this allele. Thus, 41 of the 44 alleles (93%) of PMs were defective CYP2C19m1. One of the remaining three PM alleles was subsequently found to contain the CYP2C19m2 mutation, which has earlier been shown to be associated with the PM phenotype in Oriental populations. In conclusion, the phenotype determined by omeprazole correlated with that of mephenytoin, and was in good agreement with the genotype.