The anemia of prematurity is defined by a progressive decline in hemoglobin level occurring over the first 2 months of life. Unlike term newborns whose "physiologic anemia" rarely if ever necessitates any treatment, preterm infants may become anemic enough to have clinical symptoms that indicate a need for red blood cell transfusions. Various factors contribute to the development of this anemia. Some of these factors, such as the short life span of erythrocytes in preterm infants, increased sensitivity of the erythrocytes to oxidative injury, and the blood losses caused by repeated phlebotomies, would normally be expected to induce corrective reticulocytosis. Characteristically, however, this anemia is hyporegenerative. Thus, it is associated with relative reticulocytopenia, low serum erythropoietin levels, and bone marrow erythroid hypoplasia. The recent availability of recombinant human erythropoietin has opened new perspectives in the management of a variety of anemias. Based on current knowledge of the regulation and pathophysiology of fetal and neonatal erythropoiesis, recombinant erythropoietin may represent a logical and efficient alternative to giving red blood cell transfusions in the treatment of the anemia of prematurity. Clinical trials have been initiated in several countries using different approaches and methodology. At this early stage these trials do not yet fully affirm that recombinant erythropoietin can be used as the first-line therapy in infants with the anemia of prematurity. Our own observations, however, suggest that this agent is well tolerated by preterm infants and may exert a corrective effect on the anemia of prematurity.