Short-term bioassays such as the Ames test are used to assess genotoxicity and potential carcinogenicity of specific drinking water contaminants as well as concentrated samples of drinking water. The authors discuss the development, limitations, and interpretation of short-term tests; qualitative and quantitative aspects of the utility of the tests for predicting carcinogenicity; and general approaches to using the tests in analyzing potential health effects of drinking water. They conclude that although uncertainties regarding interpretation limit the application of such tests for risk assessment, short-term tests provide public health officials with a useful tool for obtaining timely and cost-efficient information about potential health risks associated with drinking water. Includes 47 references, tables.