Radon has long been known to contribute to risk of lung cancer, especially in undergound miners who are exposed to large amounts of the carcinogen. Recently, however, lower amounts of radon present in living areas have been suggested as an important cause of lung cancer. In an effort to clarify the relationship of low amounts of radon with lung cancer risk, we placed alpha-track radon detectors in the homes of 308 women with newly diagnosed lung cancer and 356 randomly selected female control subjects of similar age. Measurements were taken after 1 year. All study participants were part of the general population of Shenyang, People's Republic of China, an industrial city in the northeast part of the country that has one of the world's highest rates of lung cancer in women. The median time of residence in the homes was 24 years. The median household radon level was 2.3 pCi/L of air; 20% of the levels were greater than 4 pCi/L. Radon levels tended to be higher in single-story houses or on the first floor of multiple-story dwellings, and they were also higher in houses with increased levels of indoor air pollution from coal-burning stoves. However, the levels were not higher in homes of women who developed lung cancer than in homes of controls, nor did lung cancer risk increase with increasing radon level. No association between radon and lung cancer was observed regardless of cigarette-smoking status, except for a nonsignificant trend among heavy smokers. No positive associations of lung cancer cell type with radon were observed, except for a nonsignificant excess risk of small cell cancers among the more heavily exposed residents. Our data suggest that projections from surveys of miners exposed to high radon levels may have overestimated the overall risks of lung cancer associated with levels typically seen in homes in this Chinese city. However, further studies in other population groups are needed to clarify the carcinogenic potential of indoor radon.