One of the most difficult decisions faced by clinicians treating myeloproliferative neoplasms (MPN), is whether or not a patient with primary myelofibrosis (PMF) should be offered an allogeneic stem cell transplant (allo-SCT). This is based on the knowledge that, although the overall median survival from the time of diagnosis is only 6 years, individual survival may vary from 1 to over 30 years, with some patients requiring little or no treatment.The therapeutic dilemma derives from the fact that standard therapies, including treatments for anemia, as well as chemotherapy and splenectomy, do not significantly affect the natural history of the disease. Furthermore, although the elucidation of the pathogenetic role of JAK2 mutations has opened up the exciting prospect of targeted therapy-currently, several JAK2 inhibitors are under evaluation and some have been shown to have significant biological activity-it is not clear if this approach will result in a survival benefit. In contrast, allo-SCT is potentially a curative procedure, although its wide-spread use is hampered by significant nonrelapse mortality and morbidity. As a result, reduced-intensity allo-SCT (RIC allo-SCT) is currently preferred for the older patient, although even this approach is not without considerable risk. Indeed, the most recent European Bone Marrow Transplant Registry (EBMTR) data, derived from 103 patients treated with a RIC allo-SCT, has shown a nonrelapse mortality of 16% at one year, with acute and chronic graft-verus-host complications of 27 and 43%, respectively [1]. Consequently, only patients with a poor outlook should be considered for bone marrow transplantation and the ''holy grail'' has been to devise a practical schema to enable such individuals to be identified accurately.A wide range of clinical and laboratory parameters have been reported by different groups to be predictive of an adverse prognosis, including advanced age, hepatomegaly, splenomegaly, constitutional symptoms, abnormal karyotype, anaemia, thrombocytopenia, low reticulocyte count, raised blood CD34 1 stem cell count, the presence of a monocytosis, or circulating blasts, and the detection of the JAK2 V617F mutation. Many of these observations, however, have not stood the test of time and only anaemia has been a universal finding. Based on a selection of the above variables, several prognostic scoring systems have been devised. The most widely used has been the ''Lille score,'' a simple schema based on just two parameters, hemoglobin <10 g/dl and leukocyte count <4 or >30 3 10 9 /l, that identifies low-, intermediate-and high-risk groups, with median survivals of 93, 26, and 13 months, respectively [2]. However, although a practical schema, it suffers from a number of limitations; firstly, the findings were based on a relatively small number of cases, indeed only 16 patients defined the high-risk group, secondly only a limited number of young patients were included in the study and lastly, the intermediate-and high-risk groups are not clearly separated. Cervan...