We conducted a case-control study on 114 bladder cancer patients and 46 hospital controls. DNA adducts were measured in WBCs by 32P postlabeling and showed no association with smoking habits and the glutathione-S-transferase M1 genotype. A strong association between adduct levels and the N-acetyltransferase (NAT2) genotype was found (P = 0.0002). The NAT2 genotype was associated in a nonstatistically significant way to the case-control status (odds ratio, 1.6; 95% confidence interval, 0.8-3.2). In a logistic regression model, the log of DNA adduct levels was associated in a highly significant way to the risk of bladder cancer (regression coefficient, 0.75; P = 0.0006), independently of smoking habits. Using the median of DNA adducts (RAL, 0.3) as a cutoff point, the odds ratio for the risk of bladder cancer was 4.1 (age-adjusted; 95% confidence interval, 1.9-9.0). Our study suggests that sources other than tobacco smoke contribute to the formation of aromatic DNA adducts in WBCs. The role of WBC-DNA adducts in predicting bladder cancer is still to be clarified.