To evaluate the association of indoor radon exposure with lung cancer risk, yearlong alpha track detector measurements of radon were conducted in dwellings which had been occupied for at least 10 years by 433 New Jersey female lung cancer cases and 402 controls who were subjects in a larger population-based study. Adjusted odds ratios were 1.1 (90% confidence interval, 0.79-1.7), 1.3 (90% confidence interval, 0.62-2.9), and 4.2 (90% confidence interval, 0.99-17.5) for exposures of 1.0-1.9, 2.0-3.9, and 4.0-11.3 pCi/liter, respectively, relative to exposures of less than 1.0 pCi/liter, showing a significant trend (1-sided P = 0.04) with increasing radon concentration. The trend was strongest among light smokers (less than 15 cigarettes/day, 1-sided P = 0.01). The trend for lung cancer risk with estimated cumulative radon exposure was slightly weaker (1-sided P = 0.09). The increase in relative risk for each unit of cumulative exposure, 3.4% (90% confidence interval 0.0-8.0%) per working level month, was consistent with the range of 0.5-4.0% per working level month generally reported for underground miner studies, supporting the extrapolation of the occupational data to the residential setting. However, the possibility of selection biases, the small number of high exposures, and other uncertainties necessitate caution in interpretation of these data.