Treatment of HCV-related chronic hepatitis is controversial when non-organ specific autoantibodies are present, due to potential severe autoimmune reactions under interferon. We evaluated, in an open study, a sequential approach (steroid->interferon) in 20 consecutive patients with biopsy-proven chronic hepatitis, anti-HCV positive (EIA2/RIBA2) and autoantibody positive at a titre > or = 1/80 (18 antinuclear and 2 anti-liver-kidney microsomal antibodies). Nine patients responded to steroids (ALT reduced by > or = 50% at 12 weeks) and continued on prednisone up to one year. Notably, ALT did not return to normal and steroid treatment was ineffective in controlling necroinflammation on follow-up biopsies. After stopping prednisone, ALT rebounded to pre-treatment levels in 6/9 cases. Four of these 6 then received interferon: 3 of them had a complete response (e.g. normal ALT at end of therapy), in 2 with loss of HCV RNA. Eleven patients were, instead, steroid resistant and after wash-out were switched to lymphoblastoid alfa-interferon (6 MU t.i.w. for 8 weeks, 3 MU t.i.w. for 16 weeks). Four cases had a complete response to interferon (3 with loss of HCV RNA) with follow-up biopsies showing definite reduction of necroinflammation. None of the 15 receiving interferon in the present study experienced ALT peaks, deterioration of liver disease, autoimmune-like phenomena. We suggest that antiviral treatment with alfa-interferon could be the first choice in chronic hepatitis C, even in autoantibody positive cases.