The mortality rate in diabetics with chronic kidney disease (CKD) is seven times higher than end-stage renal disease mainly because of cardiac causes. Anaemia may have a relevant role in the pathogenesis of cardiovascular (CV) disease in CKD. Anaemia occurs at an earlier stage of CKD in diabetic individuals than in those with other causes of CKD. Observational findings support the unfavourable influence of anaemia on mortality in CKD patients, and the combination of anaemia and CKD in diabetics identifies a group with a particularly high mortality risk. While the effect of erythropoietin on these patients' quality of life is known, its impact on mortality and CV risk is uncertain. The recent Anaemia Correction in Diabetes (ACORD) trial in diabetic CKD patients, which targeted haemoglobin levels of 13-15 mg/dl, disclosed no statistically significant favourable or adverse effects on mortality or morbidity over the 2-year follow-up, while other studies endeavouring to nearly normalize haemoglobin have reportedly proved risky. Even if anaemia is causally involved, the pathogenesis of CV disease in diabetics with CKD is so complex that addressing just one factor (anaemia) may not suffice to prevent CV risk, and normalizing haemoglobin levels may even be harmful.