Data from 11 studies of radon (10043922) (Rn) involving 65,000 exposed miners were pooled to conduct a comprehensive analysis and to develop models for the estimation of Rn associated lung cancer risk. Relative risks (RR) were calculated from working level months (WLM). A linear relationship was observed between RR and WLM, with significant RRs seen for all categories above 50WLM. Factors which decreased the WLM exposure response relationship (ERR/WLM) were increasing age, time since last exposure, and time since exposure. Duration of exposure increased the ERR/WLM. This inverse exposure rate effect indicated that for equal total exposure, risks were greater for low exposures of long duration than for high exposures of short duration. The ERR/WLM for smokers and never smokers were 0.0034 and 0.0103, respectively. A greater RR for Rn exposure was observed for never smokers than smokers, with the exposure response trend for never smokers threefold the trend for smokers. The models used for assessing lung cancer risk included cumulative exposure, time since exposure (TSE), attained age (AGE), and either exposure duration (DUR) or exposure rate. Results were presented for the TSE/AGE/DUR model. Lifetime RRs of lung cancer were estimated for female smokers to be 1.4 and 1.7 at 4 and 8 picocuries/liter, respectively. The corresponding attributable risk of lung cancer for male and females were 10% and 12% for smokers and 28% and 31% for nonsmokers. The authors conclude that 10% of all lung cancer deaths in the United States were attributable to indoor Rn, while 40% of lung cancer deaths among miners were attributed to Rn, including 70% among nonsmokers and 39% among smokers.