In a randomized prospective controlled study in humans, the metabolism and hepatic effects of a single administration of halothane were compared with enflurane and meperidine. Pre- and postoperative antipyrine pharmacokinetics, intraoperative indocyanine green clearance, liver histology, and postoperative liver function tests were determined in 24 patients undergoing abdominal surgery who were randomly allocated to receive either halothane (0.5%, group I), enflurane (0.8%, group II), or meperidine (group III), as a supplement to a common basal anesthetic regimen consisting of thiopental, nitrous oxide/oxygen/muscle relaxant. In addition, end-tidal concentrations of the volatile reductive metabolites of halothane, chlorodifluoroethylene (CDF), and chlorotrifluoroethane (CTF) were determined in group I patients and serum and urinary inorganic fluoride were determined in both group I and II patients. Indocyanine green clearance was measured before anesthesia (stage I), during basal anesthesia (stage II), in the presence of surgical stimuli (stage III), and after introduction of the selected anesthetic agent (stage IV), CDF and CTF were detectable within 20 min of the start of halothane anesthesia in every patient receiving halothane. Peak serum fluoride concentrations occurred at 2 and 24 hr in the enflurane and halothane groups, respectively, whereas urinary fluoride excretion was elevated postanesthesia in the enflurane group only. There was no difference between the pre- and postoperative disposition of antipyrine in group II or III, but after anesthesia, antipyrine clearance was significantly decreased (P < 0.02) and plasma half-life increased (P < 0.05) in group I patients (halothane). Concentrations of serum alanine aminotransferase (ALT) and bilirubin were significantly elevated (P < 0.05) postoperatively in groups I and II but unchanged from preoperative values in group III patients. Three of the 24 liver biopsies taken at the end of stage IV showed several foci of acute liver cell necrosis; of these, two patients were from group I and one from group II. There were no significant differences in liver cell morphology (P > 0.5) in biopsies taken at the end of stage IV compared with biopsies at the end of stage III, from groups I and II. The results of this study show that reductive metabolism of halothane occurs routinely in patients undergoing halothane anesthesia under conditions of normoxia. This may be the cause of the changes in antipyrine clearance after halothane anesthesia.