Initially, immune (idiopathic) thrombocytopenic purpura (ITP) was considered to be a disease of increased platelet destruction. Consequently, treatments for ITP patients were focused on reducing the destruction of platelets. However, recent research on the pathogenesis of ITP points to a role of suboptimal platelet production, and this has led to the development of new treatment options which increase platelet production such as the novel thrombopoiesis-stimulating agent, romiplostim. In two, multicentre, randomized, placebo-controlled, double-blinded, phase III trials, romiplostim was able to increase and sustain platelet counts in both splenectomized and non-splenectomized patients. Eighty-three per cent (69/83) of the romiplostim-treated patients achieved an overall platelet response compared with 7% (3/42) of patients receiving placebo (P < 0.0001; Cochran–Mantel–Haenszel test controlled for splenectomy and baseline concurrent ITP therapy). Patients from the romiplostim studies, with platelet counts below 50 × 109/L, were able to enter a long-term, open-label, extension study. The results of the long-term study showed that platelet counts above 50 × 109/L were maintained for ≥10, ≥25 and ≥52 consecutive weeks by 78% (102/131), 54% (66/122) and 35% (29/84) of patients, respectively. Romiplostim appears to be generally well tolerated and several patients have now been treated for more than 3 yrs. Most adverse events tend to be mild to moderate in nature. Romiplostim provides a valid treatment option that may bring about a change in the way that patients with ITP are treated in the future. Further studies are now being undertaken in paediatric patients with ITP, patients with myelodysplastic syndrome and patients with chemotherapy-induced thrombocytopenia.