To investigate the relationship between CYP2C19 genotypes and the hydroxylation index (HI) of omeprazole in the South Indian population. Healthy unrelated South Indian subjects (n=300) were separated into three groups based on their CYP2C19 genotypes. They were administered a single oral dose of 20 mg omeprazole, and venous blood was collected 3 h later. Plasma was assayed using reversed-phase high-performance liquid chromatography, and the omeprazole HI was calculated. The means of HIs in individuals with CYP2C19*1/*1 (n=124), *1/*2 (n=129) and *2/*2,*2/*3 (n=47) were 2.4, 5.3 and 22.5, respectively, and were found to be significantly different between any two groups (P<0.0001). A good correlation was established between CYP2C19 genotype and omeprazole HI (r=0.54, 95% CI 0.45-0.62; P<0.0001). Of the 300 subjects, 42 (14.0%; 95% CI 10.1-17.9) were phenotypic poor metabolizers (PMs), but only 33 of them had two mutant alleles and the remaining 9 PMs had at least one wild-type allele. Among the 258 extensive metabolizers, 14 had two mutant alleles. The prevalence of PMs in the South Indian population was 14.0%, which is similar to that in North Indians and Orientals but significantly higher than in Caucasians and Africans. A genotype-phenotype relationship was established between the CYP2C19 genotype and HI of omeprazole, but 7.7% of subjects deviated from expected genotype-phenotype associations. This could be due to an additional mutation, either in the exons/introns or in the 5'-regulatory region of the CYP2C19 gene.