Abstract
A major objective of studying populations exposed to low levels of radiation is provision of direct assessment of the adequacy of estimates of health risks obtained by extrapolation of data on populations exposed at high levels. Meeting this objective requires the use of statistical methods that fully utilize available exposure measurements, and that consider the extensive body of prior knowledge regarding radiation‐induced health effects. This paper describes various methods of analysis and illustrates the methods with data on workers at the Hanford Site. Special problems are discussed, with attention to comparability of estimates and confidence limits with those obtained from extrapolation.