Thymosin α1 (Tα) is an immune modifier that has been shown in a pilot study to be effective for chronic hepatitis B; this requires confirmation. Ninety-eight patients with clinicopathologically proven chronic hepatitis B were randomly allocated to 3 groups: 1) group A received a 26-week course of Tα with a 1.6-mg subcutaneous injection two times a week (T6 group); 2) group B received the same regimen as group A, but Tα therapy extended for 52 weeks (T12 group); and 3) group C served as a control group and was followed up for 18 months without specific treatment (T0 group). The three groups were comparable in clinicohistological features at entry. The complete virological response rate (clearance of serum hepatitis B virus [HBV] DNA and hepatitis B e antigen [HBeAg]) was higher in group A (40.6%) and group B (26.5%) than in group C (9.4%) (group A vs. group C: P = .004; group B vs. group C: P = .068) when assessed 18 months after entry, although complete response rates among these three groups were similar when first assessed at the end of therapy. There was a trend for complete virological response to increase or accumulate gradually after the end of Tα therapy. None of the responders lost hepatitis B surface antigen. Blinded histological assessment showed a significant improvement in treated patients, particularly in lobular necroinflammation and scores excluding fibrosis. No significant side effects were observed. These results suggest that a 26-week course of Tα therapy is effective and safe in patients with chronic hepatitis B.