In order to evaluate the clinical implication of experimental studies on halothane-induced liver damage in phenobarbital-treated rats, we studied the clinical records of 315 consecutive patients who underwent brain surgery with halothane anesthesia. After exclusion of subjects with a history of alcoholism or antecedent chronic liver disease, clinical data of 279 patients with normal preoperative transaminase activities were analyzed. The incidence of halothane-induced liver injury was significantly higher in the subjects given phenobarbital than in those with no phenobarbital medication (7/100 vs. 1/179, p less than 0.01). To determine if other anticonvulsant compounds can influence halothane-induced liver injury, rats were pretreated with diphenylhydantoin or valproic acid prior to exposure to halothane under hypoxic conditions for comparison with phenobarbital. The degree of halothane hepatotoxicity assessed from ALT activities and morphological alterations was of the decreasing order of phenobarbital greater than controls = diphenylhydantoin greater than valproic acid, and a similar order was observed in the extent of reductive metabolism of halothane. These results indicate that patients pretreated with phenobarbital may be at a greater risk of halothane-induced liver damage, and that treatment with valproic acid and diphenylhydantoin lead to the production of toxic intermediates of halothane to a lesser extent than treatment with phenobarbital does.