In chronic hepatitis C virus (HCV) infection, the rate of sustained response to interferon is low. We evaluated, in patients responding to a 26-week course of interferon, the effect of high-dose maintenance therapy in preventing relapse. Three hundred and ten patients with chronic HCV infection (38.3% with cirrhosis, 80.6% with HCV type 1) received interferon alfa-2b for 26 weeks (10 MU tiw for 8 weeks, then 5 MU tiw for 18 weeks). One hundred and twenty-four subjects (40%) normalized aminotransferases, and were allocated randomly either to continue on 5 MU tiw for a further 26 weeks (prolonged therapy group: 60 patients) or to stop interferon (brief therapy group: 64 patients). Fifty-two weeks after stopping interferon the overall sustained biochemical response rate was 13.2% (41/310). The number of patients with normal aminotransferases was comparable between the prolonged and brief therapy groups (30% vs. 35.9%, P = n.s.), and the rate of HCV-RNA clearance was similar (48.8% vs. 42.4%, P = n.s.). The timing of posttreatment relapse was not influenced by the duration of therapy. Fifty-nine patients (19%) did not complete therapy due to adverse effects. Multivariate analysis identified four features predicting sustained biochemical response in subjects normalizing aminotransferases under therapy: negative HCV-RNA at end of therapy, normal aminotransferases at 4 weeks of therapy, high baseline aminotransferases, and high baseline platelets. Infection with HCV type 1 was not a significant predictor of response, due to its high prevalence in our population (80.6%). It is concluded that in patients with chronic hepatitis C mostly infected by HCV type 1, a prolonged high-dose interferon course (900 MU over 52 weeks) did not increase the rate of sustained biochemical response and of HCV-RNA clearance in comparison to a brief course (510 MU over 26 weeks).