For uranium miners and mill workers, the main radiological risk is of
developing a cancer of the lung due to the inhalation of radon progeny
. This risk is generally expressed in terms of excess risk per unit ex
posure of airborne radon decay products (Working Level Month, or WLM),
assuming a non-threshold linear, or quadratic-linear, relationship be
tween exposure and risk. However, there are wide differences in the ri
sk factors derived from eleven major cohorts of miners exposed to rado
n progeny and, in these cohorts, the risk seems to decrease as the fre
quency of radon progeny sampling, and the reliability of exposure data
, increase. Further, in two cohorts with relatively reliable exposure
data, and in two other cohorts, the risk of lung cancer appears to be
insensitive to increasing exposure, up to about 200 to 400 WLM of tota
l exposure, a pattern which is indicative of the existence of a thresh
old for the induction of lung cancer by radon progeny. This pattern is
also consistent with other human epidemiological data and animal expe
riments which also seem to support the notion of thresholds in the ind
uction of cancer by internally deposited alpha emitters. The second pa
rt of the paper examines the difficulties in the evaluation of exposur
es and radiation doses from airborne radon progeny and long-lived radi
oactive dust. It is shown, in particular, that the evalution of doses
due to the inhalation of long-lived radioactive dust from dust samples
collected on filters requires the development of new, appropriate tec
hniques and procedures.