In an effort to meet the need for a fast and cheap in vivo prescreen for inherited mammalian point mutations, a somatic forward-mutation method, originally developed in an X-ray experiment, has more recently been tested in work with chemical mutagens. The method makes use of coat-color mutations because (a) the gene product is usually locally expressed, (b) mosaics can be detected with minimal effort, and (c) opportunities for making comparison with induction of germinal point mutations are greatest.--Following treatment of embryos that are heterozygous at specific coat-color loci, various induced genetic changes can result in expression of the recessive (RS) in clones derived from "mutant" melanocyte precursor cells. However, other events, such as decrease in the number of precursor cells, or disturbed differentiation, can also result in spots, which with careful classification can usually be distinguished from RS's on the basis of their location and color. When this is done, the relative RS frequencies for a series of compounds at least roughly parallel the relative spermatogonial mutation rates. The fact that easily measurable (though low) RS rates are obtained with compounds that have yielded negative results in spermatogonial tests is not surprising in view of the fact that RS's can be caused by several mechanisms besides point mutation.--In spite of the parallelism observed in one laboratory, the usefulness of the in vivo somatic mutation method as a prescreen could come to be doubted because of major discrepancies between results of similar experiments at different laboratories. However, it appears probable that at least some of these discrepancies are due to failure to discriminate between spots that probably resulted from melanocyte insufficiency and spots that resulted from expression of the recessive.--Reverse somatic mutation systems can potentially avoid some of the pitfalls of forward mutation systems. Such system are still in developmental stages.