Little is known of the predictive value of the levels of DNA adducts in terms of cancer risk. We examined the association between bulky DNA adducts and risk of lung cancer in a population-based cohort, comprising of 25,717 men and 27,972 women aged 50-64 years at entry. We included 245 cases (137 men and 108 women) with lung cancer and a comparison group of 255 individuals (137 men and 118 women), matched on sex, age and smoking duration. Bulky adducts in white blood cells collected at enrollment and stored at -150 degrees C were analyzed by (32)P-postlabeling method, using the butanol enrichment procedure. The median level of bulky DNA adducts was 0.196 adduct/10(8) nucleotides (5-95 percentiles: 0.094-0.595) among current smokers who were later diagnosed with lung cancer and 0.163 adduct/10(8) nucleotides (5-95 percentiles: 0.091-0.455) among current smokers in the comparison group. The smoking adjusted incidence rate ratios (IRR) for lung cancer in relation to one log unit (natural logarithm) difference in adduct levels were 1.22 (95% CI 0.85-1.74), 1.33 (95% CI 0.89-1.98) and 0.76 (95% CI 0.39-1.47) among all, current and former smokers, respectively. Current smokers with bulky DNA adduct levels above the median had a significant higher lung cancer rate than those with adduct levels below the median (IRR = 1.61; 95% CI 1.04-2.49). The results are compatible with previous studies, suggesting a slightly higher risk of lung cancer with higher levels of adducts among smokers. Our results indicate that bulky DNA adducts may have a weak association with lung cancer risk.