Other studies have indicated that hypoxia enhances the binding of halothane metabolites to components of the hepatic microsomal fraction. The authors pretreated Wistar rats with phenobarbital, 75 mg/kg, daily for four days to induce the hepatic drug-metabolizing enzyme system and subsequently made them hypoxic (FI(O2) = .08) while they were receiving halothane, 0.6 percent. Centrilobular hepatic necrosis was well developed by six hours following exposure, and early stages of resolution were evident by 48 hours. Hemorrhage occurred within the necrotic areas, and leukocytosis was not prominent. Normal rats or those depleted of hepatic glutathione did not experience hepatic necrosis when made hypoxic and given halothane. Rats receiving halothane in adequate oxygen (FI(O2) = 0.50) after phenobarbital pretreatment showed no hepatic necrosis. Plasma fluoride values were normal (< 2 μM) immediately upon completion of halothane exposure when the animals received adequate oxygen, whether or not they had received phenobarbital pretreatment. When halothane was administered to hypoxic animals, plasma fluoride values averaged 19 ± 2 μM (Mean ± SEM), and pretreatment with phenobarbital caused a further increase in plasma fluoride to 24 ± 2 μM. Plasma fluoride values thus increased indicate that the production of defluorinated halothane metabolites is caused primarily by hypoxia. In animals showing hepatic necrosis, cytochrome P-450 was decreased but cytochrome b5 was not changed. The selective decrease in cytochrome P-450 suggests a specific, persistent involvement of this enzyme rather than a generalized destruction of microsomal enzymes. It is concluded that halothane anesthesia causes hepatic necrosis in rats when combined with phenobarbital stimulation of the hepatic drug-metabolizing enzyme system and hypoxia.