To evaluate the efficacy of a 12-month course of recombinant interferon alpha (IFN-alpha2b), and to assess predictive factors of successful response to IFN therapy in chronic active hepatitis C (HCV CAH), 242 patients with histologically proven HCV CAH were assigned randomly to two groups, one treated with IFN-alpha2b (3 MU three times weekly, intramuscularly), the other untreated. To determine the efficacy of IFN-alpha2b 12 months after therapy, a second liver biopsy was carried out on 100 treated patients and 27 untreated patients. The biochemical, virological, and serological response of patients followed up for at least 50 months after treatment was also evaluated to confirm the efficacy of IFN-alpha2b. The genotypes of infecting HCV, anti-HCV core IgM, and HCV-RNA concentrations were also analysed and the predictors of response determined by univariate and multivariate analyses. Response was defined in terms of the normalisation of aminotransferase activities and the disappearance of HCV-RNA. The overall long-term response was 39.4%. Anti-HCV core IgM levels were significantly lower in long-term responders. Patients with increased levels of IgM anti HCV core (>3.8 sample/cut-off), infected with genotype 1b were nonresponders. Liver histology improved significantly in patients with long-term response. Multivariate analysis identified three independent predictors of the likelihood of long-term response to IFN therapy: age younger than 40 years, basal anti-HCV core IgM levels < or = 3.8, and genotypes other than 1b. These data indicate that the treatment with IFN-alpha2b used in this randomised controlled trial is effective in HCV CAH. Anti-HCV core IgM was the strongest predictor of long-term response in the present study.