Anaemia is an independent predictor of adverse outcomes in chronic kidney disease (CKD). Randomized trials using erythropoiesis-stimulating agents (ESAs) in patients with severe anaemia (baseline haemoglobin level <100 g/l) have been small, and the hypothesis that partial correction of severe anaemia may prevent cardiovascular events is tenable but unproven. Results from randomized trials of moderate anaemia correction with ESAs do not support the hypothesis that moderate anaemia is a cardiovascular risk factor. This Perspectives article discusses the idea that this finding may have been a result of the inadequate design of trials, co-intervention with high doses of ESAs or intravenous iron hiding a beneficial effect. Another idea is also discussed-that moderate anaemia is a marker of the degree of renal impairment and that its association with cardiovascular disease merely signifies that other factors are present that are pathogenic to the heart and associated with both kidney impairment and anaemia.