The possible mechanisms of naphthoquinone-induced toxicity to isolated hepatocytes were investigated using three structurally-related naphthoquinones, 1,4-naphthoquinone (1,4-NQ), 2-methyl-1,4-naphthoquinone (2-Me-1,4-NQ) and 2,3-dimethyl-1, 4-naphthoquinone (2,3-diMe-1,4-NQ). 1,4-NQ was more toxic than 2-Me-1,4-NQ whereas 2,3-diMe-1,4-NQ did not cause cell death at the solubility-limited concentrations used. All three naphthoquinones extensively depleted intracellular glutathione (GSH). However, the depletion of GSH induced by 1,4-NQ and 2-Me-1,4-NQ prior to cell death was more rapid and extensive than that induced by the nontoxic 2,3-diMe-1,4-NQ. Further studies demonstrated that 2,3-diMe-1,4-NQ was cytotoxic in the presence of dicoumarol, a compound which also potentiates the cytotoxicity of 1,4-NQ and 2-Me-1,4-NQ. To investigate the differential cytotoxicity of these three naphthoquinones, their relative capacities to redox cycle and to bind covalently to cellular nucleophiles were assessed. Redox cycling was investigated using rat liver microsomes where the order of potency for quinone-stimulated redox cycling was 1,4-NQ approximately 2-Me-1,4-NQ much greater than 2,3-diMe-1,4-NQ as indicated by nonstoichiometric amounts of NADPH oxidation and O2 consumption. NADPH-cytochrome P-450 reductase was implicated as the enzyme primarily responsible for naphthoquinone-stimulated redox cycling. The reactivity of the naphthoquinones with glutathione and, by implication, with other cellular nucleophiles was 1,4-NQ greater than 2-Me-1,4-NQ much greater than greater than 2,3-diMe-1,4-NQ. Overall, these studies indicate that 2,3-diMe-1,4-NQ is not cytotoxic (except in the presence of dicoumarol) and this lack of toxicity may be related either to its lesser capacity to redox cycle and/or its inability to react directly with cellular nucleophiles.