Summary. This study was designed to investigate the value of liver biopsy in the management of patients with chronic hepatitis C virus infection and to identify risk factors for fibrosis. It was a prospective audit of clinical, biochemical, virological and radiological features for predicting liver fibrosis in 140 consecutive patients seropositive for antibody against hepatitis C virus. Seventy-five per cent of patients were asymptomatic and 69% had no clinical signs of chronic liver disease. Serum transaminase levels were normal in 44% of patients, less than twice normal in 35% and more than twice normal in 21%. Ultrasound scan was unremarkable in 85%. Sixty-nine per cent of patients were viraemic at the time of liver biopsy. Liver histology revealed that fibrosis was absent in only 10% of patients (stage 0) while cirrhosis was evident in 7% (stage 5). Liver fibrosis was detected in 90% of patients (stage 1, 11%; stage 2, 41%; stage 3, 21%; and stage 4, 10%). Univariate analysis showed that increasing age, clinical signs of chronic liver disease and abnormal ultrasound scan findings were associated with liver fibrosis (P < 0.05); however, multiple linear regression analysis of all the clinical features did not reveal a useful model (sensitivity 42% and specificity 23%) for predicting liver fibrosis. Hence, no combination of clinical, biochemical, virological or radiological data was reliable in discriminating the stage of liver fibrosis in patients with chronic hepatitis C virus infection. However, older patients, especially those with clinical signs of chronic liver disease and abnormal ultrasound scan findings, were more likely to have advanced fibrosis. We recommend liver biopsy as the single most important investigation in detecting liver fibrosis.