Immune thrombocytopenia (ITP) is an autoimmune disorder characterized by low platelet counts resulting from both immune-mediated platelet destruction and inappropriate bone marrow platelet production. Therefore, in patients with ITP failing immunosuppressants/splenectomy, an alternative approach is to enhance platelet production stimulating thrombopoiesis. Studies on the development of recombinant thrombopoietins (rhTPO) were halted as a minority of patients developed an autoantibody that neutralized pegylated rhTPO and also cross-reacted with and neutralized endogenous TPO resulting in thrombocytopenia. Clinical use of romiplostim, a second-generation TPO-RAs, has shown that during long-term treatment, it may elicit the development of neutralizing antibodies to this agent resulting in acute thrombocytopenia. In our case series of 47 primary adult patients with ITP treated with romiplostim, 28 of 47 are evaluable for response loss. Among these, we observed eight patients who either progressively (3 of 8) or abruptly (5 of 8) lost response which accounts for a prevalence of 28.5%. Neutralizing antibody testing could be performed in 4 of 8 patients and 3 of 4 tested positive. These antibodies did not cross-react with endogenous TPO and retesting of 2 patients at 9 and 7 months yielded a negative result. At follow-up, 5 of 8 patients – including the 3 patients with neutralizing antibodies – went into long-term complete response when switched to a different therapy while 3 of 8 patients never regained a response on subsequent lines of therapy. Response loss does not seem to be so rare an event during romiplostim administration (28.5% in our series) and in a minority of patients, it can be associated with development of drug neutralizing antibodies. Although recognized by the manufacturer as a possible adverse event ensuing during romiplostim administration, development of neutralizing antibody in everyday clinical practice has so far not been specifically addressed in reports on romiplostim use outside controlled studies. Unfortunately, testing for these antibodies requires adhesion to strict procedures which is not easily accomplished in everyday clinical practice. This complexity represents a significant drawback in extending antibody testing to all patients who lose response to romiplostim.