Follow-up of a cohort of 1,165 steelworkers exposed to acid mists has been extended from 1981 to early 1986 for most cohort members, and information on smoking has also been collected. We obtained an SMR for lung cancer (unadjusted for smoking) of 1.56 (95% CI 1.12-2.11, 41 observed). For those with 20 years or more since first exposure, the SMR was 1.72 (1.21-2.39). However, no trend was found with duration of exposure. To adjust for smoking, we used an indirect adjustment as suggested by Axelson to account for the fact that the exposed cohort smoked slightly more than the U.S. referent population. The SMR for the whole cohort dropped to 1.36 (0.97-1.84), while for those with more than 20 years since first exposure, the SMR was 1.50 (1.05-2.07). These results indicate that an excess risk for lung cancer persists after control for confounding by smoking.