While a significant fraction of the 70,000 chemicals in commercial use in the United States are probably mutagens and/or carcinogens, only about 3,000 have been adequately tested in at least one assay. Since some of these chemicals represent a significant environmental health risk to man in terms of increased cancer and genetic disease, it is important that they be identified and characterized with the ultimate aim of limiting human exposure. To this end, many distinct assays have been developed, but each has certain inherent limitations. We recognize the need to obtain a large data base on the specific types and frequencies of mutations induced by a given substance in mammalian cells. We further recognize that these data should be obtained from different regions of a functioning mammalian gene within cell types that are representative of the specific cell types in vivo in which mutation and/or carcinogenesis occur. Thus, we have been developing an in vitro human cell assay which both detects mutagens and determine the frequency at which they cause specific transitions, transversions or frameshifts in transfected, site-specific mutated, mouse adenine phosphoribosyltransferase (APRT) genes. This gene was chosen because of its sensitive selectable characteristics. The specifically mutated mouse APRT genes can, in principle, be introduced into differentiated human cells which exhibit a spectrum of mutagen metabolizing enzymes similar to that observed in vivo. In addition, we have been developing immunological methods for directly detecting, in situ, mutagens that induce frameshifts.