Aims: To investigate the incidence of the CYP2C19 polymorphism in the Chinese Dai population. Methods: One hundred and ninety-three healthy Chinese Dai volunteers were identified with respect to CYP2C19 by genotype and phenotype analyses. A polymerase chain reaction-restriction fragment length polymorphism method was performed for genotyping procedures. The 4¢-hydroxymephenytoin (4¢-OH-MP) and S/R-mephenytoin (S/R-MP) excreted in the urine were determined by high-performance liquid chromatography and gas chromatography, respectively. Results: Eighteen subjects were identified as poor metabolisers (PMs). The frequency of PMs in the Chinese Dai subjects was 9.3% (95% confidence interval 5.2, 13.4), which is lowerthan that in the Chinese Han population (P<0.05). Chinese Dai subjects had a higher frequency of the mutant CYP2C19*2 allele (0.303) and a lowerfr equency of the mutant CYP2C19*3 allele (0.034). These two mutant alleles could explain all deficiencies of CYP2C19 activity in the Chinese Dai subjects. The frequency of the CYP2C19*3 allele is significantly lowerthan that in the Chinese Han population (P<0.05). The mean S/R ratio was lower in the homozygous extensive metabolisers (EMs) compared with that in heterozygous EMs (P<0.01), and the latter was lowerthan that in the PMs (P<0.01). Furthermore, the mean S/R ratio in CYP2C19*3/CYP2C19*2 heterozygous PMs was possibly lower than that in the CYP2C19*2/CYP2C19*2 homozygous PMs (P<0.05). Conclusion: The frequencies of PMs and CYP2C19*3 allele in the Chinese Dai population are significantly lowerthan those in the Han population. The CYP2C19 genotype analysis is largely consistent with the mephenytoin phenotype analysis. The variability of S/R ratios in EMs and PMs shows a gene–dosage effect.