Detection of abnormal liver function tests in a patient receiving an antidepressant or other psychopharmacological agent presents a major problem to the clinician, who must decide whether or not to withdraw a drug which may be of central importance in the patient's management. Experimental studies have clearly demonstrated a hepatotoxic potential for many of these compounds, but in the clinical situation the picture is far less clear cut because there are remarkably little published data on the incidence of liver damage from these agents in man. Furthermore, interpretation of abnormal liver function tests and ascribing them to a drug rather than to an alternative cause is not always straightforward. Hepatic drug reactions fall into two categories; those which are predictable and dose-related (e.g. acute hepatic necrosis following suicidal overdose with the analgesic paracetamol), and those which are unpredictable and dose-independent, complicating treatment with a drug at therapeutic dosage (Davis and Williams, 1985). Most hepatic drug reactions, including those caused by antidepressants, fall into the second category. The spectrum of severity may vary widely, from minor abnormalities in biochemical liver function tests with no constitutional disturbance, to severe hepatitis and/or cholestasis which can sometimes be fatal. Chronic liver disease has also been described.