As part of a retrospective epidemiologic study of lung cancer in women, alpha-track and charcoal canister radon measurements were made in a sample of New Jersey residences. The alpha-track measurements were designed to yield estimates of annual average exposures of dwelling occupants, while charcoal canister measurements were designed to yield "worst case" concentrations for screening purposes. The year-long living area and basement screening measurements had geometric means of 19 Bq m-3 (0.52 pCi L-1) and 56 Bq m-3 (1.5 pCi L-1), respectively. Measurements of radon gas with different detectors and on different floors were compared to each other within residences. Ratios of screening to annual average results became more extreme as the measured concentrations increased; the mean ratio of basement canisters to year-long living area alpha-track detectors was 5.6 vs. 2.7 for houses that screened above and below 150 Bq m-3 (4 pCi L-1), respectively. Although the residence sample from which these data were drawn is not necessarily representative of either state or national housing stock, these observations, if verified, may have important implications for procedures and decision strategies designed to reduce individual and population exposures to radon.