Despite the exclusion of commercial donors and the introduction of sensitive radioimmunoassays for hepatitis B surface antigen, approximately 8 to 10 per cent of prospectively followed patients continue to develop post-transfusion hepatitis. Serological evaluation of these residual hepatitis cases showed them to be distinct from hepatitis B, hepatitis A, cytomegalovirus and the Epstein-Barr virus. These cases have been tentatively designated non-A, non-B and now compromise approximately 90 per cent of post-transfusion hepatitis. There is increasing evidence that non-A, non-B hepatitis is also spread by routes other than blood. In the absence of a specific serological test, the diagnosis of non-A, non-B hepatitis depends on the clinical exclusion of non-viral causes of hepatocellular injury and on the serological exclusion of other known hepatitis viruses. The clinical pattern of non-A, non-B hepatitis is characterized as follows: (1) The incubation period tends to be intermediate between type A and type B hepatitis, but considerable overlap exists. Most cases occur between five and 12 weeks after percutaneous exposure: (2) Non-A, non-B hepatitis tends to be less acutely severe than type B hepatitis; two thirds to three quarters of cases are anicteric, demonstrate only mild to moderate transaminase elevations and are symptomatically mild. Individual cases may, however, be clinically and biochemically severe and cannot be distinguished from acute type A or type B hepatitis. (3) A striking feature of non-A, non-B hepatitis is its propensity to progress to chronic liver disease. Up to 50 per cent of patients have elevated transaminase values in excess of one year and the majority of such patients, when biopsied, demonstrate histological features of chronic active hepatitis. Approximately 10 per cent show features of cirrhosis. Despite these histological changes, the chronic liver disease of non-A, non-B hepatitis seems, in most cases, to be slowly resolving rather than rapidly progressing. Non-A, non-B hepatitis has now been repeatedly induced in chimpanzees using human acute and chronic phase inocula. These studies establish the existence of a transmissible agent and of a chronic asymptomatic carrier state. Transmission appears to be very similar to that for type B hepatitis. Additional studies in humans and chimpanzees suggest that there may be more than one non-A, non-B agent. At present, there is no confirmed serological test for detecting this agent(s). The development of such a test would represent a major breakthrough in the prevention of post-transfusion hepatitis and in better defining the epidemiological pattern and clinical consequences of this increasingly prevalent disease.