The basic premises that guide genetic toxicity testing for identifying carcinogens and to support administrative and regulatory decisions are: the Salmonella mutagenicity test is a necessary component of testing schemes; a chromosome aberration test is needed in addition to a gene mutation test; a mammalian cell mutagenicity test is needed in addition to the Salmonella test; in vivo tests are needed to confirm the results of in vitro tests; and test batteries are more predictive than the individual tests of the battery. Results from the Salmonella mutagenicity, in vitro chromosome aberration, mutations in mouse lymphoma cells, rodent bone marrow micronucleus, and rodent carcinogenicity tests, performed by the U.S. National Toxicology Program, were used to evaluate these premises. A positive Salmonella test was most predictive of carcinogenicity. However, the data do not support using the other tests in addition to Salmonella for predicting carcinogenicity. The genetic toxicity tests did not complement each other, and batteries or combinations of the tests were no more predictive of carcinogenicity than Salmonella alone. If a chemical is mutagenic in Salmonella it should be considered a potential rodent carcinogen, unless ancillary information suggests otherwise. Positive responses in the other in vitro or in vivo tests do not increase the probability that the chemical is a carcinogen, and negative responses in the other tests do not diminish the implications of the positive Salmonella response.