The French cohortof uranium miners includes 5098 miners,and 125 lung cancer deaths have been observed. The relationship between radon exposure and risk of lung cancer is estimated using relative risk models, which allow investigation of time dependent modifying factors such as the period of exposure, time since exposure and exposure rate. A linear exposure-response relationship is observed. The main modifier is the period of exposure before or after 1956: the excess relative risk after introduction of ventilation in the French mines is eight times higher than before. The decrease in risk with time since exposure and exposure rate disappears when period of exposure is taken into account. Compared to most of the uranium miner studies, this cohort constitutes a population exposed to low levels and low rates of radon for a long duration. The extension of the French cohort leads to an important increase in its statistical power. Exposure rate effect will be further investigated, in the framework of a European collaborative research project aiming at the synthesis of the effects of radon exposure at low dose and low dose rate.