[en] More than 2000 rats were exposed to cumulative doses of up to 28,000 WLMs of radon gas. More than 300 pulmonary tumors were induced by this exposure, most being nonfatal lesions detected only at autopsy of animals that had died of unrelated causes. Above 6000 WLMs rats suffered increasingly from life shortening due to radiation-induced nonneoplastic causes and so had less time in which to develop tumors. When adjusted for these competing causes of death, the hazard function for the excess risk of developing pulmonary tumors was approximately linearly related to dose throughout the range of doses studied. This suggests that some previously reported high-dose ''reductions'' in radiogenic tumor-induction rates may chiefly have involved the killing of rats rather than the killing of precursor cells. Rats exposed to radon and then to six months of inhalation of tobacco smoke had a four times greater age-specific prevalence of pulmonary tumors than rats exposed to an identical radon dose either alone or preceded by tobacco smoke inhalation. This suggests that tobacco smoke may accelerate the carcinogenic process by acting as a promoter of radiation-induced somatic damage. These data suggest that, for assessing human risk from exposure to radon, the linear model should be assumed, but that the WLM is not on its own an adequate index of carcinogenic insult. 7 refs., 2 figs., 4 tabs