Hepatic halothane extraction was studied in seven miniature swine. During halothane anesthesia, nylon catheters with Silastic tips were implanted in the aorta, portal vein and common hepatic vein. After three to five weeks halothane was given. An alveolar concentration of 1–1.3 per cent was maintained for an hour; thereafter, the concentration was decreased. The halothane fraction removed from hepatic blood flow was calculated from blood halothane concentrations. From these values it was determined that alveolar concentrations below 0.0026 per cent resulted in nearly complete removal of halothane by the liver, but no significant removal was detected at anesthetic concentrations. The fraction removed was defined as halothane metabolized. Data from this study indicate that results of tracer studies of total body uptake cannot be extrapolated to the anesthetic state. Tracer studies may cause two-to fivefold overestimates of uptake during maintenance of anesthesia, compared with investigations conducted during anesthesia. Similarly, studies of hepatic metabolism during anesthesia grossly underestimate hepatic biotransformation and production of metabolites at subanesthetic halothane levels.