Interferon alfa (IFN-α) is the primary treatment for chronic hepatitis B. The standard duration of IFN-α therapy is considered 16 weeks; however, the optimal treatment length is still poorly defined. We evaluated the efficacy and acceptability of prolonged IFN-α treatment in patients with chronic hepatitis B. To investigate whether treatment prolongation could enhance the rate of hepatitis B e antigen (HBeAg) seroconversion, we conducted a prospective, controlled, multicenter trial in which all patients were treated with a standard regimen of 10 million units IFN-α 3 times per week over 16 weeks. Patients who were still HBeAg-positive after 16 weeks of therapy were randomized to prolongation of the identical regimen up to 32 weeks (prolonged therapy) or discontinuation of treatment (standard therapy). Among the 162 patients who entered the study, 27 (17%) were HBeAg-negative after the first 16 weeks of treatment, and 118 were randomized to standard or prolonged therapy. After randomization, a response (HBeAg seroconversion and sustained hepatitis B virus [HBV]-DNA negativity) was observed in 7 of the 57 (12%) patients assigned to standard therapy versus 17 of the 61 (28%) patients assigned to prolonged therapy (P = .04). A low level of viral replication after 16 weeks of treatment, as indicated by serum HBV-DNA values under 10 pg/mL, was found to be the only independent predictor of response (52% vs. 0%; P < .001) during prolonged therapy. The prolonged IFN-α schedule was well tolerated in the large majority of patients. In chronic hepatitis B, prolongation of IFN-α therapy up to 32 weeks is superior to a standard course of 16 weeks. Those patients who exhibit a low level of viral replication at the end of the standard regimen benefit most from prolonged treatment.