The prevalence of hepatitis-C virus (HCV) infection was investigated in a group of children with chronic post-transfusion hepatitis using a first- and second-generation HCV-antibody ELISA, 2 confirmatory tests (a second-generation recombinant immunoblot assay and a line immunoassay) as well as an HCV-polymerase chain reaction (PCR). In 33% of the children, clear discrepancies were observed between the 4 different HCV-antibody detection assays, indicating that the serological diagnosis of HCV infection is still problematic. HCV RNA was detectable by PCR in only 69% of the antibody positive patients, which may be due to a fluctuation of viraemia during the course of infection. Such a fluctuation was demonstrated in 6 patients from whom serum samples drawn at different times were investigated. In contrast, in 8 of the 15 seronegative patients, HCV infection was identified only by PCR, although the hepatitis had already persisted for more than 2 years. Antibody assays and PCR together detected HCV infection in about 90% of the patients with chronic hepatitis. When markers of hepatitis B infection were also investigated, only 6% of the cases remained undiagnosed.