The existence of a rare syndrome of "enflurane hepatitis" similar to that described for halothane and of a cross-sensitization between halothane and enflurane has been controversial, largely due to equivocal clinical case reports and a lack of a plausible molecular mechanism for the hepatotoxicity. The present study suggests a possible hypersensitivity basis for enflurane hepatitis and the apparent cross-sensitization between halothane and enflurane involving covalently bound liver microsomal adducts. Immunoblotting studies have revealed that antibodies in the sera of six patients with halothane hepatitis recognize liver microsomal antigens of Mr = 100,000, or both 100,000 and 76,000, formed in rats treated with enflurane or halothane. These antigens were not detected in microsomes from isoflurane- or sesame oil-treated rats. The recognition of these antigens could be abolished by preincubation of the sera with microsomes from halothane-treated rats. These data suggest that the difluoromethoxydifluoroacetyl halide metabolite of enflurane, as well as the trifluoroacetyl halide metabolite of halothane, covalently bind to similar hepatic proteins, and may become immunogens in susceptible patients. This mechanism may also account for the apparent cross-sensitization between halothane and enflurane anesthesia, and the development of hepatic necrosis.