To assess the utility of DNA adducts as biomarkers of exposure to carcinogens in an industrial population, a pilot study of roofers occupationally exposed to a mixture of polycyclic aromatic hydrocarbons was conducted. DNA was isolated from peripheral white blood cells of roofers and non-occupationally exposed subjects matched for age, sex and smoking status. Occupational exposures to anthracene, fluoranthene, pyrene, benzanthracene, benzo[a]pyrene, benzo[b]fluoranthene, benzo[g,h,i]perylene and benzo[k]fluoranthene were assessed by personal breathing zone air sampling and skin wipes. Exposures to benzo[a]pyrene in air of exposed subjects ranged from 0.60 microgram/m3 to 1.39 micrograms/m3, and exposures to total polycyclic aromatic hydrocarbon (the sum of eight hydrocarbons) ranged from 6.0 micrograms/m3 to 13.8 micrograms/m3 on the day before blood collection. In the biomarker studies 10 of 12 roofers, but only 2 of 12 comparison subjects, had detectable levels of aromatic DNA adducts by 32P-postlabelling assay (p less than 0.01). The two non-roofers with detectable adducts had levels at or near the detection limit of 2 adducts per 10(9) nucleotides. In two roofer samples which were studied in a mixing experiment, the major adduct spots did not co-migrate with the guanosine N2 adduct of benzo[a]pyrene diol epoxide. These results suggest that the 32P-postlabelling assay may be useful for monitoring exposures to complex mixtures of aromatic hydrocarbons in industrial populations.