Monolayer cultures of rat hepatocytes activated tris(2,3-dibromopropyl)phosphate (Tris-BP) more efficiently than 2-acetylaminofluorene (AAF), to genotoxic products which caused mutations in co-cultures of S. typhimurium. In contrast, AAF caused a greater genotoxic response in the hepatocytes than Tris-BP, as judged by the increase in DNA-repair synthesis measured by liquid scintillation counting of 3H-TdR incorporated into DNA isolated from the nuclei of the hepatocytes. Covalent binding of 0.05 mM 3H-Tris-BP to cellular proteins occurred at a similar rate as covalent binding of 0.25 mM 14C-AAF. Tris-BP was the more cytotoxic of the two compounds as determined by leakage of cellular lactate dehydrogenase into the culture medium. The observed differences in the cytotoxic and genotoxic responses between Tris-BP and AAF were probably caused by differences in the nature of their reactive metabolites with respect to stability, lipophilicity and/or their interactions with variuos cellular nucleophilic sites. The relative DNA-repair synthesis induced by an AAF exposure for 18 h decreased with time after plating of isolated hepatocytes. Tris-BP first caused an increase in the relative DNA-repair synthesis up to 27 h after plating, whereafter the response declined reaching control values using cultures 75 h after plating. In parallel with the decreased relative response in DNA-repair synthesis with time, the background radioactivity in isolated nuclei from untreated cells increased both when the hepatocytes were incubated in the presence or absence of hydroxyurea to inhibit replicative DNA synthesis. Increased DNA-repair synthesis was demonstrated as early as 3 h after commencing exposure to the test substances. While the induced DNA-repair synthesis caused by Tris-BP remained constant after 6 h of exposure, the response caused by AAF increased with increased exposure time beyond 6 h. To assess the role of different metabolic pathways in the genotoxic and cytotoxic responses of Tris-BP and AAF, the hepatocytes were exposed to test substances in the presence of various metabolic inhibitors for 3 h, whereafter the cell medium was removed and replaced by cell-culture medium containing 3H-TdR and hydroxyurea. The cytochrome P-450 inhibitor metyrapone decreased both the genotoxic and cytotoxic effects of Tris-BP, while α-naphthoflavone reduced the genotoxic effect of AAF. The addition of glutathione (GSH) or N-acetylcysteine decreased both the cytotoxic and genotoxic effects of Tris-BP, while cellular depletion of GSH by diethylmaleate increased these effects. Manipulations in the cellular levels of sulhydryl-containing substances in the hepatocytes by these agents had little effects on the DNA-repair synthesis caused by AAF. The results indicate that such a hepatocyte culture system may be very useful as a tool to study mechanisms involved in the formation of cytotoxic and/or genotoxic metabolites from various xenobiotics.