The test for sex-linked recessive lethals (SLRL) in Drosophila melanogaster has been used to detect induced mutations since 1927. The advantage of the test for both screening and hazard evaluation is its objectivity in testing for transmissible mutations in the germ cells of a eukaryote. Statistical criteria for both positive and negative mutagenicity at the highest concentration tested under a particular exposure condition were developed by the Work Group, and a recommended protocol for future testing was agreed upon. For 421 compounds there were sufficient data available in the literature for analysis; 198 compounds were found to be positive and 46 negative at the highest concentration tested. Most experiments had been done for objectives of pure research rather than for deliberately screening for mutagenicity, although many of the 421 chemicals were selected for testing because of suspected mutagenicity. Therefore, the statement of 198 positive and 46 negative should not be taken as an example of the proportion of mutagens in the environment. In three sets of experiments with D. melanogaster that were done specifically for screening, one involving 40 compounds for the Environmental Protection Agency (EPA), the others involving 13 for the Food and Drug Administration (FDA), only 6 mutagens were discovered. After completion of the classification of compounds according to their response in the SLRL test, the compounds were classified as to their carcinogenic response according to the list of Griesemer and Cueto (1980). There were 62 compounds that could be classified as positive or negative for both carcinogenesis and mutagenesis. Of the 62 compounds, there was agreement between the carcinogenesis and mutagenesis classification in 56 (50 positive and 6 negative), or 90% would have been correctly classified as to carcinogenesis from only the SLRL test. Because of inadequate sample size, 177 compounds could not be classified as positive or negative according to the statistical criteria established by the Work Group. This large number of inadequately tested compounds reflects the fact that many of the experiments were not done for screening. Further work is needed on the compounds with inadequate sample size.(ABSTRACT TRUNCATED AT 400 WORDS)