A safety analysis of pooled data from clinical studies of romiplostim, a thrombopoietin (TPO) receptor agonist, in which patients with immune thrombocytopaenia (ITP) received romiplostim, placebo, or medical standard of care (SOC) Rodeghiero et al. (Eur J Haematol 91:423-436, 2013), has been updated. Included are data from 14 trials spanning 2002-2011; placebo- and SOC-arm data are pooled. Most patients (n = 1059) were female (61 %) and Caucasian (85 %); 38 % had undergone splenectomy; 23 were children. Mean (SD) baseline platelet count was 20.6 (16.5) × 10(9)/L. Mean (SD) weekly dose of romiplostim was 4.2 (2.8) µg/kg; total exposure was 1520 patient-years. Overall, 921 patients received romiplostim only, 65 received placebo/SOC only, and 73 received placebo/SOC followed by romiplostim. Rates of haemorrhage (romiplostim, 205/100 patient-years; placebo/SOC, 263/100), thrombosis (both, 5.5/100 patient-years), haematological malignancy/myelodysplastic syndrome (romiplostim, 0.5/100 patient-years; placebo/SOC, 2.7/100), and non-haematological tumours (romiplostim, 2.2/100 patient-years; placebo/SOC, 3.6/100) were comparable among groups. Bone marrow reticulin was reported in 17 patients and collagen in one patient receiving romiplostim; one patient receiving placebo/SOC had reticulin reported. Three patients developed neutralizing antibodies to romiplostim, but not to endogenous TPO. This integrated analysis of the safety profile of romiplostim in patients with ITP is consistent with previously reported studies; no new safety concerns emerged.