Cigarette smoking is an established cause of bladder cancer. The direct relationship between smoking-induced DNA adducts in bladder cells and cancer risk at that site has, however, been poorly assessed. We therefore investigated the relationship between bladder cancer risk and levels of DNA adducts measured in normal bladder biopsies by 32P-post-labeling in a hospital-based case-control study of 59 bladder cancer patients and 45 controls submitted to surgery for prostatic hyperplasia or urinary incontinence. An approximately 2-fold risk for bladder cancer was found in individuals with an adduct level >14.8 (median among controls) compared with those with an adduct level < or =14.8 (OR = 1.9, 95% CI 0.8-4.3, P = 0.13). A dose-response relationship was also suggested (trend test, P = 0.13): compared with adduct levels below 13.5, the OR for bladder cancer was 1.7 (95% CI 0.6-4.6) for adduct levels between 13.5 and 18.5 and 2.2 (95% CI 0.8-6.1) for adduct levels >18.5. These findings provide some evidence that DNA adducts in bladder tissue might predict smoking-induced bladder cancer. Larger studies are still warranted to confirm these results.