We have reviewed the results of our cytogenetic studies to evaluate the usefulness of these assays in genotoxic studies. In one study, we observed unusual dose-response in lymphocyte chromosome aberration frequencies after exposure of mice to low doses of a chemical mixture (benzene, chloroprene, epichlorohydrin, and xylene). The frequency in the high dose group is lower than those of the medium and low dose groups. This reduction of genotoxicity is correlated with a significant induction of a detoxifying enzyme, glutathione-S-transferase. The data also suggest that extrapolation of effects from high to low doses for risk assessment may be erroneous. Using benzene as a model clastogen, we found that the clastogenic effect persists for a long time after termination of exposure in mice. This phenomenon is probably due to a gradual release of benzene from absorption of the chemical in body fat. In an inhalation study, we observed that chromosome aberrations are induced in mice after exposure to benzene at below the occupational exposure limit of 1 ppm. Since benzene is a ubiquitous environmental contaminant, it may interact with other environmental agents to modify its genotoxic effects. We found that a nongenotoxic drug, praziquantel, and a free radical scavenger, DMSO, can enhance or reduce respectively the clastogenic activities of benzene in mice. Both modifying agents acted by altering the metabolic pathways of benzene. In a study with rats, we showed that carcinogenic doses of benzene can induce chromosome aberrations in lymphocytes of rats long before the expression of cancer. With dibromochloropropane, we observed that this chemical can induce dominant lethality in rats.(ABSTRACT TRUNCATED AT 250 WORDS)