The association between the distribution characteristics of CYP2A6 catalytic activities toward nicotine and coumarin, and the frequency distribution of CYP2A6 variant alleles reported was estimated in 120 healthy Thais. The distributions of the subjects as classified by the amounts of 7-hydroxycoumarin (7-OHC) excreted in the urine and by cotinine/nicotine ratio in the plasma were clearly bimodal. However, the numbers of apparently poor metabolizers for coumarin and nicotine were different. The inter-individual variability in the in vivo dispositions of coumarin and nicotine closely related to the CYP2A6 genetic polymorphism. There was a close correlation between the rate of 7-OHC excretion in the urine and cotinine/nicotine ratio in the plasma among subjects (R=0.92, p<0.001). The frequency of CYP2A6 allele found in the present study was: CYP2A6*1A=32% (95% CI, 22.1-39.4%), CYP2A6*1B=27% (95% CI, 19.4-33.5%), CYP2A6*9=20% (95% CI, 17.6-23.3%), CYP2A6*4=14% (95% CI, 9.6-17.8%), CYP2A6*7=5% (95% CI, 3.7-9.4%), CYP2A6*10=2% (95% CI, 0.8-5.1%). Subjects having CYP2A6*1A/*1B were found to have a higher rate of 7-OHC excretion, as well as a higher cotinine/nicotine ratio in the plasma compared with those of the other genotypes. In contrast, subjects with CYP2A6*4/*7 and CYP2A6*7/*7 almost lacked any cotinine formation, whereas urinary 7-OHC was still detectable. CYP2A6*9 allele clearly resulted in reduced enzyme activities. Despite the absence of the homozygote for CYP2A6*10 allele, the presence of CYP2A6*10 allele significantly decreased the enzyme activities. The results of the present study demonstrate that in vivo phenotyping of CYP2A6 using nicotine and coumarin are not metabolically equivalent. Nicotine is a better probe according to its specificity, while coumarin is still valuable to be used for a routine CYP2A6 phenotyping since the test employs a non-invasive method.