Background Patients with diabetes and anemia are at high risk of cardiovascular disease. The Anemia CORrection in Diabetes (ACORD) Study aimed to investigate the effect of anemia correction on cardiac structure, function, and outcomes in patients with diabetes with anemia and early diabetic nephropathy. Methods One hundred seventy-two patients with type 1 or 2 diabetes mellitus, mild to moderate anemia, and stage 1 to 3 chronic kidney disease were randomly assigned to attain a target hemoglobin (Hb) level of either 13 to 15 g/dL (130 to 150 g/L; group 1) or 10.5 to 11.5 g/dL (105 to 115 g/L; group 2). The primary end point was change in left ventricular mass index (LVMI). Secondary end points included echocardiographic variables, renal function, quality of life, and safety. Results Median Hb level and LVMI were similar in groups 1 and 2 (Hb, 11.9 and 11.7 g/dL 119 and 117 g/L; LVMI, 113.5 and 112.3 g/m2 , respectively). At study end, Hb levels were 13.5 g/dL (135 g/L) in group 1 and 12.1 g/dL (121 g/L) in group 2 ( P < 0.001). No significant differences were observed in median LVMI at month 15 between study groups (group 1, 112.3 g/m2 ; group 2, 116.5 g/m2 ). Multivariate analysis showed a nonsignificant decrease in LVMI ( P = 0.15) in group 1 versus group 2. Anemia correction had no effect on the rate of decrease in creatinine clearance, but resulted in significantly improved quality of life in group 1 ( P = 0.04). There were no clinically relevant differences in adverse events between study groups. Conclusion In patients with diabetes with mild to moderate anemia and moderate left ventricular hypertrophy, correction to an Hb target level of 13 to 15 g/dL (130 to 150 g/L) does not decrease LVMI. However, normalization of Hb level prevented an additional increase in left ventricular hypertrophy, was safe, and improved quality of life.