The lymphocyte bulky PAH-DNA adduct levels have been studied in persons occupationally exposed to ambient air pollution. The exposure group consisted of 90 healthy, nonsmoking bus drivers from the Copenhagen area, divided into three exposure groups according to driving area, and 60 rural controls (smokers and non-smokers). PAH-DNA adducts were determined by 32P-postlabelling with the butanol enrichment procedure. The bus drivers answered a comprehensive questionnaire on passive smoking, residential area, diet and other potential confounding variables. A significantly higher adduct level was observed in bus drivers working in central Copenhagen (1.214 fmol/microg DNA, n = 49) compared with both those driving in the dormitory (median: 0.507 fmol/microg DNA, P = 0.046, n = 16) and suburban (median: 0.585 fmol/microg DNA, P = 0.041, n = 25) areas. All three groups had higher adduct levels than rural controls (0.074 fmol/microg DNA, n = 60, P < 0.001). No significant influence on adduct levels was demonstrated from potential confounders, including smoking and diet. The effect of the metabolizing enzymes, GSTM1 and NAT2, on adduct levels was investigated. No statistically significant effects were observed on adduct levels from GSTM1 or NAT2, either individually or combined, but a non-significant trend was seen for individuals with GSTM1*0/0 (null), since they had higher adduct levels in all exposure groups. This study demonstrated that lymphocyte PAH-DNA adduct levels were related to levels of exposure to urban air pollution and indicated that these adducts might be helpful as a means of classifying better different exposure groups for epidemiological studies. Furthermore, it demonstrated the ability of 32P-postlabelling to discern small differences in low exposure to ambient air pollution and suggested a possible effect of GSTM1*0/0 on DNA adduct levels.