Investigations of cancer and cytogenetic damage among reinforced-plastics workers have produced contradictory results. In all studies, the focus has been on styrene rather than the carcinogen, styrene-7,8-oxide (SO), traces of which are generated during the manufacturing process. Because styrene is present at very high levels and is metabolized almost exclusively through SO, coexposures to SO have been discounted. This study investigated the relative contributions of airborne styrene and SO and of smoking toward several SO-specific biomarkers (DNA and albumin adducts) and sister chromatid exchanges in the blood of 48 reinforced-plastics workers. Among individual subjects, albumin and DNA adducts as well as sister chromatid exchanges were significantly correlated with styrene exposure. However, among the 20 subjects with measurements to both styrene and SO, albumin adducts were significantly correlated with exposure to SO but not to styrene. Finally, among the 10 job groups, surprisingly strong correlations (0.709 < or = r < or = 0.966) were found between all SO biomarkers and exposure to SO but not to styrene. Calculations suggest that SO was about 2000 times more effective than styrene in producing SO biomarkers. After accounting for the disparate exposures to the two chemicals, a typical worker received 71% of the systemic dose of SO via inhalation; nonetheless, 5 of the 20 subjects received the majority of the SO dose from styrene. Cigarette smoking increased levels of SO-albumin and SO-DNA adducts, suggesting that SO was a constituent of tobacco smoke. We conclude that inhalation of SO should be considered in any interventions to reduce health risks.