To compare the long-term effect of natural lymphoblastoid interferon-alpha (IFN-alpha nl) and recombinant IFN-alpha 2a therapy in patients with chronic hepatitis B, 210 patients in two trials were followed-up for 1.1-15.5 years following the end of therapy. They included 34 patients who received placebo (control), 67 treated with IFN-alpha nl (36 after prednisolone priming) and 109 treated with IFN-alpha 2a (56 after prednisolone priming). The cumulative sustained response was higher in patients who had been treated with IFN-alpha nl after prednisolone priming than was exhibited using IFN-alpha nl alone, IFN-alpha 2a alone or the placebo (P < 0.05), or IFN-alpha 2a following prednisolone priming (P = 0.052) at the end of 11 years. Hepatocellular carcinoma (HCC) was detected in 1.5% of the IFN-alpha nl group, 3.7% of the IFN-alpha 2a group and 14.7% of the control group (control vs IFN-alpha nl or IFN-alpha 2a, P < 0.05). The cumulative HCC development was higher in the control group than in the IFN-alpha nl group (P < 0.002) and the IFN-alpha 2a group (P = 0.06). The cumulative survival rate was lower in the control group than in the IFN-alpha nl group (P < 0.01) and the IFN-alpha 2a group (P = 0.02). Multivariate analysis revealed that IFN-alpha nl therapy and female gender are significant predictors of sustained response; preexisting cirrhosis, age at entry and IFN therapy are significant factors in both HCC development and survival. In conclusion, IFN-alpha nl treatment may have a better long-term effect on hepatitis B virus (HBV) clearance than IFN-alpha 2a and placebo, and IFN therapy may provide better long-term beneficial effects than placebo in terms of HBV clearance, reduction of HCC and prolonged survival.