Abstract： Objective To evaluate the efficacy and safety of rhTPO (recombinant human thromobopoietin) employed for the treatment of sepsis-associated thrombocytopenia. Method There were 47 patients with sepsis-associaiod thrombocytopcnia eligible for the prospective, randomized (random number) and controlled clinical study from January 2009 to November 2009 in ICU of the Tianjin First center Hospital. According to the principle of minimum distribution imbalance index, these patients were randomly divided into the rhTPO group (n = 21) and the IVIG (intravenous immunoglobulin) control group (n = 22). In the rhTPO group, rhTPO was given subcutaneously to patients in a dose of 300 U/kg/d for 2 ～ 8 d, and in the IVIG control group, IVIG was used instead of rhTPO in a dose of 400 mg/kg/d for 5 days. Laboratory tests included blood routine examination, hepatic function, kidney function, coagulation function. The amount of blood products used, bleeding events, the days of ICU and hospital stay, total therapy cost and 28-day mortality were compared between two groups. Results The maximal platelet count in the rhTPO group was significantly higer than that in the contral group (t = 2.21, P =0.032). The mean value of difference between minimal and maximal platelet counts in the rhTPO group was much higher than that in the control group (t =7.40, P ＜0. 001). The average platelet count was no statistical difference between two groups before treatment (t =0. 458, P ＞ 0.05), but the average platelet counts in the rhTPO group were significantly higer than those in the contral group on the second and third day after treatment(t = 2. 166 and t = 2. 132, P =0. 036 and P =0.041. There were no statistical differences in incidence of bleeding, length of ICU stay and mortality between two groups (χ2 =0.720, t =0.91 and χ2 =0.264, P ＞0.05) , but the amounts of plasma and platelet transfusion were significantly less in the rhTPO group than those in the control group (t = 2.038 and t =2.252, P=0.048 and P=0.030) and the medical cost was cut down significantly in rhTPO group (t = 16.93, P ＜ 0.001). There was no adverse reaction occurred during period of observation. Conclusions The rhTPO can significantly increase platelet count, and decrease the amount of blood transfused and the medical cost. The administration of rhTPO is safe and efficient for the treatment of sepsis-associated thrombocytopenia.