To identify defective alleles of CYP2C19 (CYP2C19*2 and *3) in North Indians.One hundred extensive metabolizers and 21 poor metabolizers of omeprazole were genotyped with respect to CYP2C19*2 and *3 alleles with polymerase chain reaction-based diagnostic tests.Fifty-two extensive metabolizers and six poor metabolizers were homozygous with the CYP2C19*1/*1 genotype, and 48 extensive metabolizers and six poor metabolizers were heterozygous with the CYP2C19*1/*2 genotype. Nine poor metabolizers were homozygous with the CYP2C19*2/*2 genotype. No extensive or poor metabolizers demonstrated the presence of the CYP2C19*3 allele. CYP2C19*2 could explain 43% (9/21) of the poor metabolizers and 57% (24/42) of the defective alleles in poor metabolizers. Allele frequency of CYP2C19*1 and *2 was 0.7 (95% confidence interval of 0.65 to 0.75) and 0.3 (95% confidence interval of 0.25 to 0.35), respectively. Homozygous extensive metabolizers excreted 7.85 +/- 7.6 micromol 5-hydroxyomeprazole in 8 hours, which was 28% more as compared with heterozygous extensive metabolizers who excreted 5.6 +/- 3.6 micromol 5-hydroxyomeprazole in 8 hours (P < .05).CYP2C19*2 demonstrated allele frequency of 0.3, whereas CYP2C19*3 was absent in North Indians. Because CYP2C19*2 is not able to explain 57% of poor metabolizers, other mutations (CYP2C19*4 to *8) might be present in North Indians. CYP2C19 demonstrated differential evolution in North Indians because the frequency of CYP2C19*2 was similar to that in Oriental subjects, but that of CYP2C19*3 was similar to that in white subjects.