The advent of recombinant human erythropoietin (epoetin) in the late 1980s transformed the management of renal anaemia, liberating many dialysis patients from lifelong regular blood transfusions, in turn causing severe iron overload and human leucocyte antigen sensitization. Epoetin can be administered either intravenously or subcutaneously, but the half-life of the drug is fairly short at around 6–8 hours, necessitating frequent injections. To circumvent this problem, two manipulations to the erythropoietin molecule were engineered. The first of these was to attach an extra two carbohydrate chains to the therapeutic protein hormone (to make darbepoetin alfa), and the second was to attach a large pegylation chain to make continuous erythropoietin receptor activator. Both of these strategies prolonged the circulating half-life of the erythropoietin analogue. The next erythropoietic agent to be produced was peginesatide, a peptide-based agent which had no structural homology with native or recombinant erythropoietin, but shared the same biological and functional characteristics. Future strategies include stabilization of hypoxia-inducible factor, by orally active inhibitors of the prolyl hydroxylase enzyme, and advanced clinical trials are underway. In the meantime, several large randomized controlled trials have highlighted the potential harm in targeting a near normal haemoglobin of 13–14 g/dL (with an increased risk of cardiovascular complications), and sub-normal correction of anaemia is now advised. Some patients may show mild or severe resistance to erythropoiesis-stimulating agent (ESA) therapy, and common causes include iron insufficiency, infection, and underlying inflammation. Very rarely, patients may produce antibodies against their ESA, which neutralize not only the ESA, but also endogenous erythropoietin, causing pure red cell aplasia.