A case-control study, nested in a cohort of workers under surveillance for silicosis in 1979 or later, was undertaken to assess lung cancer risk in relation to the ILO coding scheme for the pneumoconioses. The subjects of this study are from the 41 matched quarters, consisting of one workers with silicosis and three age-matched controls, in which a lung cancer case was diagnosed. The odds ratio for lung cancer among subjects with ILO classification 1/0 or more, in comparison to subjects with category < or = 0/1, was 3.27 (95% CI =1.32-8.2). Adjustment of the radiographic risk for the effect of cumulative radon exposure had the effect of increasing the odds ratio for the association between ILO category > or = 1/0 and lung cancer. Although small smoking differences could account for the increased lung cancer odds ratio among workers with silicosis, the empirical evidence suggests that these smoking differences do not exist. It is concluded on the basis of two North American studies of silica exposed workers that radiographic silicosis is a marker for an increased risk of lung cancer.