In this study we evaluated the effects of the genetic polymorphism of aldehyde dehydrogenase2 (ALDH2) on toluene metabolism and determined biological exposure indices (BEIs) for toluene by the genotypes of ALDH2. The study subject were 77 men workers who are handling toluene in a video tape manufacturing factory and a textile company. Through the face-to-face interview, the information about smoking and drinking behavior wag obtained. For determination of ALDH2 poly morphism, 5 ml of venous blood sample was obtained from each subject after informed consent. DNA was extracted from the buffy coat and ALDH2 genotyping were performed using a polymerase chain reaction (PCR) method. The genotype of ALDH2 was classified into the homozygous genotype of normal ALDH2 (NN), and the heterozygous genotype of normal and inactive ALDH2 (ND), and homozygous genotype of an inactive ALDH2(DD). The concentration of hippuric acid (HA), the main metabolite of toluene, was determined in urine specimens collected at the end of shift, corrected with creatinine (HA/C), and compared with BEI for toluene, which is 2.5 g/g creatinine. The personal exposure level of toluene were monitored, using personal air sampler and analyzed by gas chromatography. The frequencies of the three genotypes in this study subjects were, NN : 45 (58.4%), ND : 26 (33.8%) and DD : 6 (7.8%), and frequencies of the genotypes in the middle or heavy toluene exposure workers were not significantly different from those in the mild toluene exposure workers. The frequencies of the DD type of ALDH2 was lower among alcohol drinkers than among non-drinkers. The urinary HA concentration of DD group was significantly lower than that of the NN or ND group, which suggests that the HA formation from toluene decreased in DD group. Regression lines were used to estimate the BEIs of the NN, ND, and DD groups. NN : y = 0.0085x + 0.23, r = 0.90 ND : y = 0.0074x + 0.21, r = 0.85 DD : y = 0.0041x + 0.82, r = 0.83 The three regression lines revealed that the estimated urinary HA concentration of NN, ND, and DD groups at 377 mg/m3 toluene(TLV-TWA) exposure were 3.43, 3.00, and 2.37 g/g creatinine, respectively. The HA concentration of the NN, and ND group were higher than that of the ACGIH BEI (2.5 g/g creatinine) ; however, the HA level of DD group was lower than the BEI. These results suggests that the ACGIH BEI is not adequate to estimate the toluene exposure of the NN, ND and DD groups at the same time. Based upon those results, a new BEI for ALDH2 deficient workers may be necessary.