To investigate the efficacy of using recombinant human interleukin 11 (rhIL-11) to reduce the need for platelet transfusions, we performed a randomized, double-blind phase II/III study with 110 acute myelogenous leukemia (AML) patients in the first complete remission. Following chemotherapy patients were subcutaneously administered either placebo (n=37) or rhIL-11 at a dose of 25 microg/kg (n=37) or 50 microg/kg (n=36). rhIL-11 administration was well tolerated. There was no difference between the rhIL-11 and placebo groups in the frequency and volume of platelet transfusions. In a perprotocol analysis set (101 patients), the platelet transfusion frequency in the 50-microg/kg group (3.0 +/- 1.76 times) was significantly lower than in the placebo group (3.9 +/-2.35 times; multiplicity-adjusted P= .049). We analyzed infection-related events retrospectively. The frequency of fever was significantly decreased in the 50-microg/kg, 25-microg/kg, and placebo groups (66.7%, 70.3%, and 89.2%, respectively; P= .018, Cochran-Armitage test). Stomatitis was less frequent in the 50-microg/kg and 25-microg/kg groups (2.8% and 0%, respectively) than in the placebo group (21.6%, P= .0012). These results show that rhIL-11 does not reduce the platelet transfusion requirement in AML patients, but the retrospective analysis confirms the previous finding that rhIL-11 reduces infection in patients undergoing chemotherapy.