In order to estimate the prevalence of HCV carriage in an inner city health district, we undertook a polymerase chain reaction (PCR) based survey of sera collected from 1,0002 patients attending general practitioners for reasons unrelated to liver disease. The series comprised 305 sample selected sera patients sample from sera 995 patients previously screened by C100 antigen-based anti-HCV tests. Overall, 7 patients were positive for HCV RNA. Four cases had anti-C100 antibodies to HCV, 2 were strictly negative but had high-normal/borderline optical densities by ELISA assay, while one was completely anti-HCV negative. All but one had normal liver function tests. Only 3/7 PCR positive cases had any serum marker for hepatitis B (HBV) exposure (2 HBsAg positive, 1 IgM anti-HBc positive). The minimum point prevalence of HCV carriage in this community is 0.7%, approximating the HBsAg carriage in the same population (1%). HCV carriage in this inner city population is considerably higher than would be predicted by blood donor surveys. A positive anti-HCV anti-body (anti-C100) test is poorly predictive (approximately 10%) of HCV RNA carriage in a general practice based population in which measurement of "surrogate" (HBV related) HCV markers would have detected only 3/7 cases of presumed chronic HCV carriage.