Serious hepatotoxicity is uncommon with the proper therapeutic use of non-narcotic analgesics but experience with new non-steroidal anti-inflammatory drugs (NSAIDs) is limited. Drugs such as ibufenac, fenclofenac and benoxaprofen were withdrawn from the market because of hepatotoxicity, and liver damage has been reported on occasion with virtually all non-narcotic analgesics. However, a clear pattern of toxicity with characteristic clinical, biochemical and histopathological abnormalities has emerged with relatively few. With the exception of acute hepatic necrosis following overdosage of paracetamol, little is known of the mechanisms of liver injury induced by non-narcotic analgesics. Involvement of the liver in a generalised drug reaction does not imply specific hepatotoxicity. About 50% of patients given aspirin regularly in anti-inflammatory doses develop mild, dose-dependent reversible liver damage as shown by elevation of the plasma aminotransferase activity. Liver damage is more severe in a small minority and it may rarely be complicated by disseminated intravascular coagulation and encephalopathy with a fatal outcome. There have also been isolated reports of chronic active hepatitis associated with the use of salicylates. Salicylate hepatitis has been reported most often in young females with connective tissue diseases. Many patients with Reye's syndrome have been given aspirin during the prodromal phase, and this serious condition closely resembles subacute salicylate intoxication in children. Salicylate probably has a causal or contributory role in Reye's syndrome, but many refuse to accept this and the issue is the subject of heated debate. Paracetamol in overdosage causes acute hepatic necrosis, and liver damage has been attributed to its therapeutic use. However, most reports have involved chronic alcoholics who took excessive doses and in these patients the clinical, biochemical and pathological findings were typical of paracetamol overdosage. Many authors have failed to make the distinction between therapeutic use and a therapeutic dose. In other cases liver damage could have been caused by exposure to other agents, viral infection or naturally occurring liver disease. If these cases are excluded, there are very few reports of liver damage associated with the proper therapeutic use of paracetamol. In some cases, the picture resembled chronic active hepatitis but no causal relationship has been established between this condition and paracetamol use. Paracetamol does not cause deterioration in liver function in patients with chronic liver disease.(ABSTRACT TRUNCATED AT 400 WORDS)