In the early years of the 21st century, results from a number of epidemiologic studies of populations with specific ionizing radiation exposures will become available. These include populations with accidental exposures in the former Soviet Union and elsewhere and populations with occupational exposures from routine operations of nuclear power plants. The strengths and limitations of these studies are reviewed together with the radiation protection questions they may answer. Many of these studies will provide specific information to complement the atomic-bomb survivor studies, particularly the effects of dose-rate and exposure protraction, modifiers of radiation risks (both environmental and host factors), and different types of radiation. These studies will therefore be important as a test of the adequacy of the current scientific bases for the radiation protection of workers and the general public. An example is thyroid cancer risk in young children following the Chernobyl accident, which has brought attention to a very high sensitivity of very young children that was difficult to assess on the basis of atomic-bomb data alone. Radiation protection will also benefit from formal comparisons and combined analyses of data from populations with different exposure patterns and exposures. Finally, future epidemiological studies will be most valuable if they are well focused, designed specifically to answer outstanding radiation protection questions. An integrated approach based on epidemiology and mechanistic studies, in which epidemiologic studies are designed to test specific mechanistic hypotheses and realistic mechanistic models are used for the analysis of epidemiological data, will probably be the most fruitful for radiation protection.