Radiation genetics has demonstrated that mutagenesis is a complex process affected by many factors. The ABCW hypothesis, that mutation frequency per rad over a wide range of organisms, from microbes to man, is linearly related to DNA content, ignores the fact that, within the mouse alone, different cell stages exhibit a range of mutation rates greater than that listed for the whole evolutionary tree. Also ignored are the findings that the important effects of dose rate and some other factors in the mouse were not predictable even from Drosophila. A much greater maze of complexities has already been found in chemical mutagenesis. This is illustrated even by the results obtained from testing of a single drug. Thus, it is clear that the attempt to extend the ABCW hypothesis to chemicals will be of little, if any, predictive value. Similarly, such concepts as the REC (roentgen-equivalent-chemical), designed to express the mutagenic risk from a chemical by a single unit quantitatively related to radiation damage, are defeated by the extreme qualitative differences in response. Unifying theories and simple non-mammalian tests that reliably predict the results in mammals cannot be expected to materialize until much more information has been collected on transmitted mutations induced in mammalian germ cells.