This multicenter, randomized trial assessed the efficacy and safety of a recombinant human thrombopoietin (rhTPO) in patients with persistent primary immune thrombocytopenia (ITP) who had failed glucocorticosteroid treatment. A total of 140 eligible patients were randomized to receive rhTPO + danazol (rhTPO group, 73 patients) or danazol (control group, 67 patients) alone. During the first phase, the increase in the mean maximal platelet counts (101.2 x 10(9)/L) and the area under curve (749.6) in the rhTPO group were significantly higher compared to control (33.3 x 10(9)/L and 316.2; P = 0.0060 and 0.0000, respectively). The major response rate (MRR) and total response rate (TRR) in the rhTPO group were 38.4 and 60.3 %, respectively, significantly higher than in control (MRR 7.9 %, P = 0.0003; TRR 36.5 %, P = 0.0104). In the control group, 45 patients with platelet counts < 20 x 10(9)/L were given rhTPO during the second phase and achieved MRR 31.1 % and TRR 66.7 %. The mean platelet counts in the rhTPO group were still approximately 50 x 10(9)/L on day 28 of the study. The overall incidence of rhTPO-related adverse events was 13.6 %. All the adverse events were generally mild. This study demonstrated that rhTPO was well tolerated, and it markedly increased platelet counts in chronic ITP patients. Stimulation of platelet production by rhTPO may provide a new therapeutic option for patients with ITP.