This is a critical review of the reports of postoperative jaundice and liver necrosis that have been published during the past twenty-five years. Postoperative liver damage has the following causes: transmitted, provoked, or coincidental viral infections; acute sepsis and septicaemic states; blood transfusion; widespread vasoconstriction with hypotension lasting more than twenty-four hours; allergic reactions to antigenic substances; alleged "hypersensitivity" to drugs to which antibodies cannot be identified; the direct effects of certain therapeutic agents on the liver; surgical trauma to the liver and its appendages; and postoperative pancreatitis. Several of these factors may operate simultaneously in the patient recovering from a surgical operation and in these circumstances it is impossible to dissociate one particular cause from the others. The syndromes of viral hepatitis and drug "hypersensitivity" are indistinguishable. Both are associated with fever, skin rashes, leucopenia, arthralgia, cholestatic jaundice, and liver necrosis with a mononuclear infiltration of the liver. Both are peculiar to man and the time interval between the exposure to risk and the appearance of jaundice varies from less than seven days to more than three months. The development of immunity to the one or desensitization from the other is not universal. There is no inhalational anaesthetic agent in common use which comes within the definitions of directly or indirectly hepatotoxic substances. The pharmacological methods of screening inhalational anaesthetic agents for hepatotoxidty have always been reliable and there is no reason to doubt their efficacy at the present time.