A case-control study was carried out among male residents in Madrid from 1982 through 1985 to ascertain occupational risk for laryngeal cancer. The study covered 50 histologically confirmed cases, 43 hospital, and 46 population controls. Occupational history and lifetime consumption patterns for cigarettes and alcohol were obtained by interview. Risk estimates, adjusted for tobacco and alcohol consumption, were calculated using nonconditional logistic regression. The highest odds ratio (OR) corresponded to woodworkers. For them, the risk increased with duration of exposure and decreased with the number of years elapsed since leaving it. The OR for woodworkers exposed over 20 years was 5.63, 95% confidence interval (CI) 1.15 to 26.64, being even greater in the furniture workers subgroup, OR 6.67 (95% CI 1.05 to 42.57). Other occupational categories with high OR were transport drivers (OR 3.31, 95% CI 0.98 to 11.22) and bricklayers and masons (OR 2.31, 95% CI 0.85 to 6.33). Wood dust or chemical compounds used in the treatment of the wood could underlie the strongest association found.