A means of assessing hepatitis NANB virus infection, via blood transfusion, as a cause of chronic liver disease was investigated in a hospital in each of two cities in England. Patients with chronic liver disease were matched for age and sex with other patients in the same hospital and histories taken included details of previous operations and blood transfusions; if these were within ten years of the study enquiries were made of hospital records officers. All positive histories were found correct, but about one third of previous transfusions had been omitted by both case and control patients. Sixty-seven male and 35 female patients with chronic liver disease and their controls were included in analyses. The only clear difference which emerged related to residence for more than a year in the Middle or Far East by male case- (40%) or control- (21%) patients (P less than 0.05). The exclusion of patients with this history left only 34 pairs in which five (15%) of the case patients and one (3%) of the control patients had a transfusion history: this difference was not statistically significant. Although the study results have shown no clear evidence of blood transfusion as a major cause of chronic liver disease in Britain, the study method, with sufficient numbers to allow analyses of newly diagnosed patients with confirmed transfusion histories, could be used to provide an ongoing assessment of the risk.