The efficacy of interferon (IFN) combined with ribavirin for the treatment of patients with hepatitis C who failed to respond to initial IFN therapy is not well established. The primary goal of this study was to perform a systematic review of the literature evaluating the efficacy of combination therapy in nonresponders. Studies were retrieved from MEDLINE, abstracts of scientific meetings, and review of the bibliographies of retrieved studies. Controlled trials were included in the primary analysis whereas uncontrolled trials and trials reported as abstracts were included for sensitivity analysis. The primary endpoints were biochemical and virologic response. A combined estimate of the odds ratio (OR) for each endpoint was obtained by using the random effects model. The number needed to treat (NNT) was calculated by taking the inverse of the pooled risk difference. Nine controlled trials (789 patients) were identified. Six months after treatment, the overall sustained biochemical and virologic responses to 24 weeks of combination therapy were 15.2% and 13.2% with a common OR of 3.8 (95% confidence interval [CI] 2.2-6.7) and 4.9 (95% CI 2.1-11.2) compared with patients treated with IFN monotherapy. The pooled risk difference for the sustained virologic response (SVR) to combination therapy was 7% (95% CI 2-13). The NNT was 14 (95% CI 8-50), suggesting that approximately 14 patients would need to be treated with 6 months of combination therapy for 1 patient to have a SVR. A number of variables were associated with a high response rate in individual studies. Sensitivity analysis of preliminary trials suggest a higher response rate with longer duration of therapy and non-type 1 genotypes.