We reviewed randomised clinical trials evaluating the effect of lymphoblastoid or recombinant alpha-interferon in non-A, non-B chronic hepatitis. The outcomes assessed were the rates of serum alanine aminotransferase normalization and relapse during and after stopping interferon. Data were pooled by meta-analysis and a 50% overall rate difference, favouring treated patients, was found. Results showed homogeneity in direction of treatment effect both after short-term (2-6 months, greater than or equal to 2 mega-units thrice weekly) and long-term (9-18 months, variable dose) interferon course. Moreover, results did not change when type of publication (abstracts vs. full reports) and treatment duration or schedule were accounted for. About 50% of patients originally responding to treatment relapsed within 6 months of either dose reduction or stopping interferon, thus suggesting that only in about one out of four patients is benefit from treatment sustained up to 1 year. We conclude that larger trials are needed to identify an optimal schedule of treatment and to evaluate predictors of interferon effectiveness in patients with non-A, non-B chronic hepatitis.