A group of patients all regularly receiving blood products for a congenital bleeding disorder were investigated for evidence of past infection with hepatitis A and B, and compared with a control group. There was a significantly higher percentage of the test group who had markers of infection with these viruses, and 15 patients had symptoms of hepatitis during the study. The greater number of the test group showing evidence of past hepatitis A infection can only really be explained by acquiring antibody from blood products, especially since none of our patients with hepatitis had hepatitis A and there is evidence that this is not an infection associated with blood products. We conclude that the majority of patients in our study showing symptoms of hepatitis had non-A, non-B hepatitis.