Unusual mutation patterns in lung tumors among underground miners have been indicated, suggesting radon-specific alterations in the genome, but the data are not consistent. To investigate the association between residential radon exposure and p53 mutations in lung tumors, we performed a study on cases from a nation-wide population-based investigation in Sweden. Our study included 83 nonsmoking lung cancer cases and 250 smoking lung cancer cases, diagnosed 1980-1984, with a time-weighted average radon exposure over 140 Bq/m3 or up to 50 Bq/m3. Radon was measured in dwellings occupied by the study subjects at some time since 1947. Information on smoking habits and other risk factors was obtained from questionnaires. After exclusions because of the initiation of treatment or insufficient material, the p53-status of 243 tumors was determined using PCR-single-stranded conformation polymorphism analysis and sequencing determination of exons 5-8. The overall mutation prevalence was 23.9%. An increased mutation prevalence was suggested among those with high exposure to residential radon [odds ratio (OR), 1.4; 95% CI, 0.7-2.6], especially among nonsmokers (OR, 3.2; 95% CI, 0.7-15.5), but no specific mutational pattern was indicated. Furthermore, the mutation prevalence seemed to be higher among smoking lung cancer cases than among nonsmoking cases (OR, 2.1; 95% CI, 0.9-5.0), and particularly among those smoking less than 10 cigarettes per day. It may be concluded that residential exposure to radon seems to contribute to a higher mutation prevalence of the p53 gene in lung tumors.