In order to study the correlation of ambient toluene and xylene exposure with their biological monitoring indices, we measured the concentration of toluene and xylene in the workplace. We also measured their biological monitoring indices of workers from August to November in 1994. The exposed group consisted of 103 male workers and the non-exposed group consisted of 34 male workers. The ambient concentration of toluene was 44.7 +/-55.4 ppm and that of xylene was 2.35+/-2.15 ppm. The urinary concentration of hippuric acid in exposed group was 1.72+/-1.53 g/g creatinine and that of non-exposed group was 0.34+/-0.28 g/g creatinine. The difference was statistically significant between two groups (p<.0.001). The urinary concentration of o-cresol in exposed group was 692.9+/-710.8 ug/g creatinine and that of non-exposed group was 184.7+/-167.6 microgram/g creatinine. The difference was also statistically significant between two groups (p<0.001). The urinary concentration of methylhippuric acid which was compensated with urinary creatinine in I exposed group was 62. 7+/-104.6 mg/g creatinine and that of non-exposed group was 64.0+91.5 mg/g creatinine. However the difference was not statistically significant between two groups (p>0.05). When compensated with urinary creatinine, the correlation coefficient of ambient toluene with urinary hippuric acid and ocresol were 0.63(p=0.0001) and 0.65(p=0.0001), respectively. When compensated with urinary specific gravity, the correlation coefficient of ambient toluene with hippuric acid and ocresol were 0.525 (p=0.0001) and 0.547 (p=0. 0001), respectively. The compensation method using urinary creatinine provided a higher correlation coefficient. We could not find any statistically significant differences between the duration of work and other variables (urinary hippuric acid, o-cresol and methylhippuric acid). In order to monitor biological indices of toluene-exposed workers, we suggest the measurement of urinary hippuric acid rather than o-cresol. As the compensation method, we suggest to use urinary creatinine rather than urinary specific gravity.