This study was designed to compare the efficacy and safety of thymosin-alphal (T-alpha1) with that of interferon-alpha (IFN-alpha) in patients with chronic hepatitis B who were positive for hepatitis B virus (HBV) DNA and hepatitis B envelope antibody (anti-HBe). Fifty-six patients were randomly divided into groups A and B. Both groups were comparable (p > 0.05) at baseline regarding age, sex, and alanine aminotransferase (ALT) levels. Group A patients received T-alpha1 1.6 mg subcutaneously twice weekly, while group B patients received IFN-alpha 5 million IU daily for 15 days, then thrice weekly for 6 months. Results from the 2 groups were compared with data from a group of 30 patients never treated with IFN-alpha and who were followed-up for 12 months (historical control [HC] group); the 3 groups were comparable (p > 0.05). After treatment, a complete response (ALT normalization and HBV DNA loss) occurred in 8 of 26 patients in group A (30.8%) and 14 of 30 in group B (46.7%; chi2 = 1.476, p = 0.224). After a follow-up period of 6 months, a complete response was observed in 11 of 26 patients in group A (42.3%) and 7 of 30 in group B (23.3%; chi2 = 2.299, p = 0.129). The rate of complete response was significantly greater in the IFN-alpha than HC group at the end of therapy (46.7% vs 3.3%; chi2 = 15.022, p = 0.0001), and in the T-alphal than HC group at the end of follow-up (42.3% vs 3.3%; chi2 = 12.566, p = 0.0001). Ten of the 12 T-alphal responders (i.e. partial responders; 83.3%) experienced sustained, non-detectable HBV DNA after 6 months' treatment; 6 of the 14 T-alphal non-responders (42.9%) showed a delayed response of non-detectable HBV DNA during the follow-up period. Corresponding values for group B patients were 50% (9/18) and 0% (0/12). The rate of delayed response was significantly higher in group A than the other 2 groups (chi2 = 6.686, p = 0.010; chi2 = 4.964, p = 0.038), whereas the rate of flare was higher in group B than in the other 2 groups (chi2 = 3.445, p = 0.063; chi2 = 7.668, p = 0.006), during the follow-up period. Unlike IFN-alpha, T-alphal was well tolerated, i.e. no adverse effects were noted in group A. These results suggest that a 6-month course of T-alpha1 therapy is effective and safe in patients with anti-HBe-positive chronic hepatitis B; T-alpha1 can reduce HBV replication in such patients. Compared with IFN-alpha, T-alpha1 is better tolerated and seems to induce a gradual and more sustained normalization of ALT and loss of HBV DNA. Combination therapy with T-alpha1 and IFN-alpha or nucleoside analogs for hepatitis B warrants further study.