A technique was developed for the assay of the genetic activities of carcinogens in both the soma and germ line in the course of early larval development and to assess the extent of their modification through the introduction into the genome of the MR IInd autosome P-type transposable elements. The influence of MR on genotoxicity for a given treatment was indicated by the relative frequencies of non-MR (Cy) to MR-carrying sibs emerging within the same cultures. The viable genetic changes in the soma were classified as recombinational or mutational events on the basis of the comparative yields of mosaic sectors in females heterozygous for the markers y w sn carried in standard order (XS) or multiply inverted (XIn) X-chromosomes. The results with this technique are here described for the carcinogen DMN. In the absence of MR, the topical application of DMN induced no larval lethality up to the highest tested doses (20 mM), but raised the yields of the somatic sectors for all the test markers in the emerging females in accordance with a linear dose fit. Comparison of the XS and XIn data indicated that somatic mutagenesis by DMN entailed the induction of recombinational and mutational events in roughly equal proportions. The introduction of MR into the genome, whether patro- or matroclinously, resulted in dramatic and proportionately equivalent enhancements in the activities of DMN, both with respect to the induction of larval lethality and somatic sectoring. These activities increased exponentially with dose, following a 4th-degree polynomial course up to 10 mM, when larval lethality approached 100%. At lower dose levels, the yields of all sector types in the viable females also followed comparable polynomial curves at different heights, except for y sn in the XIn series, where sector recovery remained at the control level throughout the examined dose range. Analysis of the sector size distribution for eye and bristle mosaicism in the XS control and DMN-treated series gave statistically comparable mean values, irrespective of the presence or absence of MR, indicating that DMN alone, or in conjunction with the P elements, did not alter the timing of aberrant clone initiation. In contrast, estimates of the genetic induction events in the somatic primordia with 5 mM DMN indicated greatly increased response in the presence of MR, which was in excess of 20-fold for the mutational changes involving the sn locus.(ABSTRACT TRUNCATED AT 400 WORDS)