Premature infants are at risk of developing significant anemia with adverse clinical outcome, yet blood transfusions used to treat anemia carry additional risks. Studies have investigated the use of recombinant erythropoietin (r-Epo) to reduce the need for blood transfusions. The beneficial impact of r-Epo use is controversial, and the original studies and meta-analyses were hampered by significant heterogeneity across trials, particularly in gestational age/birth weight, r-Epo dose, and transfusion criteria, as well as variation in the quality of evidence.