As erythropoiesis is a continuous process, we hypothesized that a new approach with smaller doses of darbepoetin-α administered more regularly could result in fewer variations of its serum levels and be more effective. In a prevalent hemodialysis population, we evaluated the effects of the darbepoetin splitting in "mini-doses". A 34-month prospective study was performed in a cohort of 110 patients, with a mean age of 66.6 ± 14.2 years, median dialysis vintage of 63.6 months (IQ range: 27 - 86), 46.4% (n = 51) male gender, 36.4% (n = 40) diabetics. In the first 12 months of the study, the usual dose regime for darbepoetin (weekly, every other week, or monthly) was unchanged. In the following 22 months, darbepoetin doses were divided into multiples of 10 µg and administered in the maximum possible number of hemodialysis (HD) sessions. Paired Student's t or Wilcoxon matched-pairs analyses were performed. After darbepoetin splitting, we verified a decrease in its consumption (1.9 ± 1.7 to 1.4 ± 1.6 µg/kg, p < 0.0001) and in the erythropoietin resistance index (8.6 to 6.7 IU/kg/week, p = 0.0001), with stable hemoglobin levels (11.4 ± 0.7 to 11.3 ± 0.6 g/dL, p = 0.02). We also observed an increase in ferritin levels (490 ± 138.2 to 560.7 ± 149.3 µg/L, p < 0.0001) and C-reactive protein (CRP) levels, with no changes in iron doses (2.2 - 3.2 mg/kg, p = 0.24). From these results, we conclude that the splitting of darbepoetin in "mini-doses" is highly effective, potentially allowing a significant decrease in the costs of anemia treatment. Larger and randomized studies are needed to prove the cost-benefit of this new strategy. .