Considerable controversy surrounds the question of whether cigarette smoking has the potential to increase the prevalence of small opacities on chest roentgenographs among asbestos-exposed workers. To compare the relative contribution of smoking with other predictors of the presence of roentgenographic small opacities, we examined 661 men enrolled in a double-blind, randomized trial designed to assess the efficacy of vitamin A and beta-carotene in the prevention of lung cancer among workers with heavy occupational asbestos exposure. Subjects in the study population had a mean latency of 35 yr from first asbestos exposure and a mean of 28 yr in their trade. The prevalence of roentgenographic abnormalities consistent with asbestos exposure was 26% for pleural abnormalities alone, 10% for parenchymal abnormalities alone, and 20% for pleural and parenchymal abnormalities together. We investigated occupation, age, latency from first asbestos exposure, and smoking status as predictors of roentgenographic small opacities. Smoking history, independent of latency, contributed to the prevalence and extent of small opacities, but its effect was less than that of latency. We conclude, that in the setting of heavy occupational exposure to asbestos, cigarette smoking confers added risk for the development of roentgenographic small opacities.