The oxidative biotransformation of 1, 2-dichlorobenzene (1, 2-DCB) was investigated using hepatic microsomes from male Wistar, Fischer-344 and Sprague-Dawley (SD) rats, phenobarbital (PB)- and isoniazid (ISO) pretreated male Wistar rats and from man. In addition, microsomes from cell lines selectively expressing one cytochrome P450 (P4502E1, 1A1, 1A2, 2B6, 2C9, 2D6, 2A6 and 3A4) were used. The rate of conversion was 0.09 nmol/min/mg, protein for both Wistar and Fischer-344 rat microsomes, 0.04 for SD-microsomes and 0.14 for human microsomes. Induction of Wistar rats with isoniazid (ISO, a P4502E1 inducer) or phenobarbital (PB, a P4502B1/2 inducer) resulted in an increased conversion rate of 0.20 and 0.42 nmol/min/mg, protein, respectively. Covalent binding of radioactivity to microsomal protein was similar for Wistar, Fischer and ISO-pretreated rats (16-17% of total metabolites), whereas induction with PB resulted in an increased covalent binding of 23% of total metabolites. Covalent binding was 31% for SD- microsomes and only 4.6% for human microsomes. Ascorbic acid notably reduced the amount of covalently bound metabolites for the SD-microsomes only, indicating that for these microsomes quinones were likely to be involved in this part of the covalent binding. Conjugation of epoxides with glutathione (GSH) inhibited most of the covalent binding for all microsomes. In the absence of GSH, the epoxides were hydrolyzed by epoxide hydrolase, resulting in the formation of dihydrodiols. Inhibition of epoxide hydrolase resulted in a decreased conversion and an increased covalent binding for all microsomes tested, indicating a role of epoxides in the covalent binding. Fischer-344 rat liver microsomes showed a lower epoxide hydrolase activity than microsomes from Wistar and Sprague-Dawley rats, which may explain the higher sensitivity to 1,2-DCB induced hepatotoxicity of Fischer rats in vivo. Conjugation of the epoxides with GSH was predominantly non-enzymatic for the rat, whereas for man, conjugation was nearly exclusively catalyzed by glutathione-S-transferases. This difference may be explained by the formation of a 'non-reactive' 3,4-epoxide by P4502E1 in human microsomes: incubations with microsomes selectively expressing human P4502E1 as well as human liver microsomes, resulted in the formation of similar amounts of 2,3- and 3,4- dichlorophenol (DCP), as well as two GSH-epoxide conjugates in equal amounts. For rat microsomes, one major GSH-epoxide conjugate was found, and a much higher covalent binding, particularly for the PB-microsomes. Therefore, we postulate that rat P4502B1/2 preferentially oxidizes the 4,5-site of 1,2- DCB, resulting in a reactive epoxide. Postulating these epoxides to be involved in the mechanism(s) of toxicity, human risk after exposure to 1,2- DCB will be overestimated when risk assessment is solely based on toxicity studies conducted in rat.