Transmitted genetic changes in mammals can be used to study all of the main endpoints of mutagenesis: point mutations, chromosome breakage, with or without rearrangement, and chromosome nondisjunction. Four methods most commonly employed in whole-mammal germ-line mutagenicity tests as well as an in vivo somatic prescreen, are summarized. Genetic basis, historical background, description of the test, limitations, and strengths are presented for each of the five systems. The specific-locus test using visible markers is the most reliable and practical method for detecting heritable point mutations, including small deficiencies, and does not require very large numbers of animals for risk extrapolations, if a relatively high dose can be administered without killing germ cells. For the detection of chromosome-breakage events, dominant lethals are useful to determine relative sensitivities of different germ-cell stages of the male, but the heritable-translocation test is more sensitive when the exposed cells are male meiotic and postmeiotic stages. Chromosome breakage events in the female are best revealed through a sex-chromosome loss test, which utilizes genetically marked X chromosomes. The same method can also be employed to detect nondisjunction in either sex. The in vivo somatic mutation test is useful as a prescreen for both point mutations and losses of chromosomal material. The reliability and efficiency of whole-mammal mutagenicity tests must be considered in two contexts: in the assessment of mutagenicity per se (as this applies to genetic changes transmitted to future generations), and in the use of mutagenicity as a possible indicator of carcinogenicity. In the former context, the whole-mammal tests are irreplaceable because they provide the closest practicable approach to the metabolic pathways existing in man, and because there is no array of lower-system tests that can predict the complexity of the response of the different mammalian germ-cell stages (which differ greatly with respect both to their absolute and their relative sensitivities to the induction of various genetic endpoints). In the second context, i.e., the use of mutagenicity as a screen for carcinogenicity in the exposed individual, most of the whole-mammal tests are of more limited utility, because they require several weeks or months for completion. Of the methods discussed, the spot test appears most suitable, because it provides a relatively rapid in vivo system capable of detecting both gene mutations and chromosomal changes of various kinds in somatic cells, including some that have been suggested to be involved in tumor promotion.