The in-vivo somatic mutation method developed by us in an earlier X-ray experiment was tested for its usefulness in chemical mutagenesis work, specifically in the prescreening for germinal point mutations. In order to explore possible parallelisms, the 7 compounds chosen for study, as well as the genetic markers used, were those with which large-scale specific-locus mutation-rate experiments in germcells had been conducted in the past or were in progress. From 1--3 dose levels were tested for each compound. On day 10 1/4 after copulation of C57BL females with T males, a single injection of the test compound was administered, and about 2000 offspring altogether were subsequently scored for survival, morphology, and presence of spots of various types. In accordance with our earlier results we found 3 types of spots: white near midline ventral spots (WMVS) which probably result from killing of melanocyte precursor cells; spots resulting from misdifferentiation; and the remainder, which probably result from expression of the recessive by one of several mechanisms (RS). Induction of teratogenic effects, which were stage-specific rather than agent-specific, generally paralleled induction of WMVS's. Both are interpreted as resulting from cell killing. Induction of RS's did not always parallel induction of WMVS's, but roughly paralleled relative frequencies of specific-locus mutations induced in spermatogonia by the same compounds. Even though the in vivo somatic-mutation method probably detects genetic changes additional to point mutations, the results indicate that it may be a useful prescreen for germinal specific-locus mutations, provided care is taken to distinguish between the 3 types of spots, only one of which (RS) is indicative of expression of the recessive.