Hematologists are under increasing pressure to prescribeiron chelation for seemingly any patient being transfused.In the week before writing this editorial, I received 10 mail-ers and emails warning me of the dangers of iron overloadfor a variety of patient types and disease states. However,in reviewing the data, it is very unclear the true risks oftransfusion-related iron overload—a necessary first stepbefore considering expensive and potentially toxic therapy.The most abundant support for iron chelation in heavilytransfused patients is in thalassemia [1]. Although transfu-sions can prolong survival, after years endocrinopathiesand cardiac failures limit life span. Cohorts of thalassemiapatients treated in the chelation era show improved survivaland decreased cardiac deaths. While cited as an examplefor other transfused-related iron overload states, severalcaveats remain. One is that death from iron toxicity in tha-lassemia starts occurring in the second decade of life—atimeframe considerable longer than most patients withacquired anemia [2,3]. Second, considerable controversyremains about the most effective form of chelation [1].Recently, there has been a remarkable increase in anxietyabout transfusion iron overload in acquired hematologic dis-ease states ranging from the bone marrow transplantpatient to patients with myelodysplasia (MDS) [4]. However,a review of this data show this apprehension seems to bebased on selective reading of the literature.One error is the equating of the poor prognosis of need-ing to receive transfusion with the toxicity of iron overload.The relationship between transfusion dependence and pooroutcome in MDS is very clear [5–7]. However, there is lackof correlation with number or duration of transfusion. Theconnection between need for transfusion and poor progno-sis is not surprising and is seen in many other diseasessuch as sickle cell anemia [8]. Indeed, for cancer, overallthe presence of anemia is a significant adverse prognosticfactor [9]. Clearly, many factors related to transfusion otherthan iron overload are responsible for the poorer prognosis.Second, there is increasing use of the serum ferritin as asurrogate for iron overload. The relationship between se-rum ferritin and total body iron stores has always been acontroversial one. Most studies correlating serum ferritinand measure iron stores have found a poor correlation [10–12]. Serum ferritin is a powerful acute phase reactant andlevels even >1000 ng/dl can be seen in inflammatorystates or in liver disease [13,14]. The touted relationshipbetween elevated serum ferritin and poor outcomes in stemcell transplant patients is more a reflection of inflammation/liver toxicity than iron overload from transfusion, especiallybecause most patients in these studies received far 20 units of blood which has been the thresh-old for chelation in many iron overload guidelines.This article by Tefferri’s group shows in another acquiredanemia—primary myelofibrosis—that need for transfusionsis a poor prognostic factor but not peak serum ferritin ortotal transfusions [19]. They studied 185 consecutivepatients evaluated at the Mayo Clinic with a median followup of 28 months. While serum ferritin correlated with trans-fusion need and age, in multivariate analysis, only transfu-sion dependence predicted shorter survival. Of addition im-portance was the finding that need for transfusions was in-dependent of prognosis determined by the newInternational Prognostic Scoring System [20]. These find-ings are consistent with their study of patients with goodprognosis MDS [5]. Together, these studies add to the evi-dence that decreased erythropoiesis—as indicated by needfor transfusions—is a powerful adverse risk factor for sur-vival in these hematological diseases. Conversely, the num-ber of transfusions or the ferritin was not prognostic inthese patients, many receiving chronic transfusions. Thesestudies also show that a group targeted for chelation—patients predicted to have years of transfusions—have noindication of iron toxicity to justify therapy.The plain, simple, unvarnished truth is that there is noevidence to justify the current hysteria over transfusion-related iron toxicity in patients receiving chronic transfusiontherapy. To defend the use of expensive therapy requiresmore than anecdotes, post hoc observations, and studiesusing inaccurate markers of iron stores. To validate the useof chelation therapy requires high-quality studies with truemeasures of tissue iron and of cardiac toxicity showing thatendpoints improve with chelation with benefit to the patient.The challenge in MDS and myelofibrosis is to avoidtransfusions by better therapies to reverse the stem celldefect. Unfortunately, the recent focus on iron overload asa priority and not the stem cell defect is leading to a misal-