In France, about 5,000 miners have been employed in uranium mining industry since 1946. Most of these miners have experienced relatively low annual exposures in comparison to other cohorts of miners. Consequently, the hypothesis to be tested in our study was the potential risk of cancer in a relatively low exposed population, characterised by a long period of underground work. A first analysis of the ''oldest'' cohort (1,785 miners having worked underground before 1972) has been published in 1993, based on a follow-up to December 1985. Recently, this follow-up has been extended up to 1994. Compared to the first analysis, the size of the cohort has increased by 24 %, and the number of lung cancer deaths has increased by about 90 % (from 45 to 85 deaths). Mean cumulated exposure to radon is of 71.5 WML, protracted over more than 15 years. After 1956, radon exposure is lower than 2 WLM/year for 50 % of the miners. The Standardised Mortality Ratio for lung cancer is 1.65, with confidence interval CI95% = [1.3-2.0]. The Excess Relative Risk coefficient for lung cancer with cumulated exposure to radon is ERR/WLM 0.40 % (p = 0.05). This estimate is very similar to the one obtained in the first analysis. The contribution of this population to the estimation of the risk coefficient of lung cancer, in relation to low dose-rates and low cumulative exposure is far from negligible. Enlargement of the French cohort exposed after 1956 is ongoing. Moreover a European joint analysis of cohorts with low levels of exposure rates (French, Czech and German cohorts) will be performed in a near future.