This case-control study nested in the French cohort of uranium miners provides an opportunity to take account of silicosis and smoking in the assessment of the relation between radon and lung cancer. The study includes 100 miners who died of lung cancer and 500 matched controls born within the same period of birth and of the same age at the time of death of the matching case. Data on radon exposure are obtained from individual monitoring of the miners, and data on smoking come from medical records and interviews. To identify cases of silicosis among the 600 miners surveyed, appraisals carried out as part of the compensation process for occupational diseases are used. Statistical analyses are based on a conditional logistic regression, and the linear model for excess relative risk was used to model the risk of death due to lung cancer according to cumulative radon exposure. The percentage of missing data on silicotic status is less than 20%. The study reveals a significant association between the relative risk of lung cancer and silicosis (ORsilicosis = 3.6; 95% CI: 1.4-8.9), and the relation between radon and lung cancer persists after adjusting for smoking and silicotic status (ERRradon per WLM = 1.0%; 95% CI: 0.1-3.5%). Radon, cigarette smoking and silicotic status appear to be three factors that each have a specific effect on the risk of lung cancer. This study reminds us of the complexity involved in assessing occupational risks in the case of multiple sources of exposure.