We have previously reported the qualitative results of a major study on 65 pesticides (Waters et al., 1982). Dose information from this investigation (either lowest effective or highest ineffective dose tested) has now been incorporated into a computerized data management system. This report focuses on the qualitative profiles of genetic activity produced by these pesticides and our efforts to classify them according to their genotoxic effects and chemical structures. Three main categories may be distinguished based on the qualitative results: Category 1 pesticides were active in most of the in vitro and in vivo assays employed. These 9 compounds include the structurally similar organophosphate insecticides, acephate, demeton, monocrotophos and trichlorfon; the phthalimide fungicide analogues, captan and folpet; and the thiocarbamate herbicide analogues, diallate, sulfallate and triallate. The 26 Category 2 compounds demonstrated fewer positive results and may be subdivided into two parts, one of which contains 12 halogenated aromatic or heterocyclic ring compounds, including the phenoxy herbicides, 2,4-D, 2,4-DB and 2,4,5-T. The remaining part of Category 2 (14 compounds) consists of structurally similar organophosphate insecticides, azinphos-methyl, crotoxyphos, disulfoton, methyl parathion; three similar ethylenebisdithiocarbamate fungicides, maneb, mancozeb, and zineb; three similar pyrethroid insecticides, allethrin, chrysanthemic acid, and ethyl chrysanthemate; and four structurally diverse compounds, cacodylic acid, dinoseb, sec.-butylamine and benomyl. The third category of 30 pesticides gave negative results in all tests and represents structurally diverse compounds. Using the computerized profile matching methodology, from 2080 possible pairwise chemical combinations of the 65 pesticides, 20 statistically significant pairs were selected, 6 groups of pesticides were identified which were substantially similar to groups of pesticides we had formed previously (Waters et al., 1982) based on genetic activity and chemical structure. The matches showed excellent qualitative and, in most cases, excellent quantitative agreement. Hence it appears that specific patterns of test results present in the genetic activity profiles are related directly to chemical structure. Conversely, the data suggests that certain groups of compounds may be recognized by a well defined series of concordant tests results. As additional data is added, comparison of test results for new chemicals with existing data for known genotoxicants should aid in the evaluation of potential genetic health hazards.