The use of darbepoetin alfa to treat anemia in patients with lower-risk myelodysplastic syndromes (MDS) was evaluated in a phase 3 trial. Eligible patients had low/intermediate-1 risk MDS, hemoglobin ⩽10 g/dl, low transfusion burden and serum erythropoietin (EPO) ⩽500 mU/ml. Patients were randomized 2:1 to receive 24 weeks of subcutaneous darbepoetin alfa 500 μg or placebo every 3 weeks (Q3W), followed by 48 weeks of open-label darbepoetin alfa. A total of 147 patients were randomized, with median hemoglobin of 9.3 (Q1:8.8, Q3:9.7) g/dl and median baseline serum EPO of 69 (Q1:36, Q3:158) mU/ml. Transfusion incidence from weeks 5-24 was significantly lower with darbepoetin alfa versus placebo (36.1% (35/97) versus 59.2% (29/49), P=0.008) and erythroid response rates increased significantly with darbepoetin alfa (14.7% (11/75 evaluable) versus 0% (0/35 evaluable), P=0.016). In the 48-week open-label period, dose frequency increased from Q3W to Q2W in 81% (102/126) of patients; this was associated with a higher hematologic improvement-erythroid response rate (34.7% (34/98)). Safety results were consistent with a previous darbepoetin alfa phase 2 MDS trial. In conclusion, 24 weeks of darbepoetin alfa Q3W significantly reduced transfusions and increased rates of erythroid response with no new safety signals in lower-risk MDS (registered as EudraCT#2009-016522-14 and NCT#01362140).