During a 29 month period, 46 patients with chronic hepatitis C virus (HCV) received recombinant human interferon alpha-2a for 6 months and were followed for another 6 months. The dose of interferon was three million units thrice weekly and was increased to six million units if amino transferase levels failed to return to normal after 2 months of therapy. At the end of the treatment 19 patients had a complete response, 6 had a near complete response, 2 patients had breakthrough during treatment, and the remaining 19 did not respond at all. Six months after treatment only 10 of the 19 responders remained in remission. Post-transfusion disease was associated with a significantly higher remission rate than sporadic disease (9/22 vs. 1/24, P < 0.001), as was also found in non-cirrhotic compared to cirrhotic patients (9/27 vs. 1/19, P < 0.001). Age, sex, duration of disease, serum aminotransferase, albumin, bilirubin, alkaline phosphatase, or Child's classification did not correlate with treatment response. Severe side effects necessitating cessation of treatment occurred in six patients, four of whom had major autoimmune phenomena. We conclude that careful selection of HCV patients with favorable response characteristics (post-transfusion etiology and non-cirrhotic liver) and without autoimmune manifestations can improve the remission rate and decrease the complication rate during interferon treatment.