A case referent study of occupational risk factors for lung cancer in white and black males was conducted. Data were taken of residents of the Detroit, Michigan area from the Occupational Cancer Incidence Surveillance Study. The cases consisted of 3,792 primary cases of lung cancer diagnosed among black and white males, 40 to 84 years old, from 1984 through 1987. The referents consisted of 1,966 cases of colorectal cancer diagnosed among black and white males during the same period. The subjects or suitable surrogates were interviewed to obtain information on demographic characteristics and lifetime work, smoking, medical, and residential histories. Possible associations between employment in occupations and industries described by 1980 United States Census Bureau classification codes and lung cancer were examined by logistic regression techniques. By racial composition, 2,866 lung cancer cases were white and 926 black. A total of 1,596 colorectal cancer cases were white and 370 were black. White males employed as concrete and terrazo finishers, grinding machine operators, heat treating machine operations, miscellaneous machine operators, truck drivers, driver sales, and laborers, black males employed as farm workers, automobile mechanics, painting machine operators, furnace operators, and trash collectors, and black and white males employed as farmers, slicers, cutting machine operators, and assemblers had significantly increased risks for lung cancer that increased with increasing duration of employment after controlling for age at diagnosis and cigarette smoking. In those occupations in which both white and black males had increased lung cancer risks, the risk was greater for black males. Different patterns in lung cancer risk were also seen when industry of occupation was considered. White males employed in railroads, coal mining, and grocery stores and black males employed in agriculture, meat production, protective services, and automobile repair had significantly increased risks for lung cancer that increased with years of employment.