In the 'doubling-dose' method currently used in genetic risk evaluation, two principle assumptions are made and these are: (1) there is proportionality between spontaneous and induced mutations and (2) the lesions that lead to spontaneous and induced mutations are essentially similar. The studies reported in this paper were directed at examining the validity of these two assumptions in Drosophila. An analysis was made of the distribution of sex-linked recessive lethals induced by MR, one of the well-studied mutator systems in Drosophila. Appropriate genetic complementation tests with 15 defined X-chromosome duplications showed that MR-induced lethals occurred at many sites along the X-chromosome (in contrast to the known locus specificity of MR-induced visible-mutations); some, but not all these sites at which recessive lethals arose in the MR-system are the same as those known to be hot-spots for X-ray-induced lethals. With in situ hybridization we were able to demonstrate that a majority of MR-induced lethals is associated with a particular mobile DNA sequence, the P-element, i.e. they arose as a result of transposition. The differences between the profiles of MR-induced and X-ray-induced recessive lethals, and the nature of MR-induced and X-ray-induced mutations, thus raise questions about the validity of the assumptions involved in the use of the 'doubling-dose' method.