Radon, an established lung carcinogen, remains the single most importa
nt environmental radiation exposure. Yet, an excess of lung cancer fro
m breathing radon in homes has not been consistently demonstrated in s
tudies conducted to date. To address several major problems that have
hindered previous studies of lung cancer and radon in homes, we have e
mbarked upon a lung cancer case-control study in Gansu Province, China
, where a substantial proportion of the population live in underground
dwellings. In this paper, we report on results of a pilot study in wh
ich radon measurements were made for 3 days in the summer in 40 homes
under normal occupancy conditions using short-term E-PERM detectors an
d for 6 months from February through August in 49 homes using long-ter
m alpha-track detectors. Useable E-PERM data were obtained from 38 hom
es and useable alpha-track data from 47 homes. For both types of detec
tors, measurements were approximately log-normally distributed. Arithm
etic and geometric means were 233 and 185 Bq m(-3) (range 74-1,590 Bq
m(-3)) for E-PERM measurements and 165 and 158 Bq m(-3) (range 74-592
Bq m(-3)) for alpha-track measurements, respectively; 68% of E-PERM me
asured homes and 55% of alpha-track measured homes exceeded 148 Bq m(-
3). Alpha-track measurements made at the entry, middle, and rear areas
of the underground dwellings did not differ significantly (arithmetic
means of 168, 162, and 165 Bq m(-3) with standard deviations 63, 73,
and 48, respectively), which suggests that air circulation may be mini
mal. The underground dwellings measured in the pilot study had high ra
don levels and the underground dwellers may provide an excellent popul
ation for studying indoor radon and risk of lung cancer.