Studies on baboons and preliminary observations in three patients with sickle cell anemia (SS) suggested that high doses of pulse administered recombinant human erythropoietin (rHuEPO) stimulate F-reticulocyte production. We now report on the administration of rHuEPO in a double-blind format to ascertain frequency of response and potential precipitation of side effects. Ten patients were enrolled, but one was discontinued due to the indication of a blood transfusion. Of the other nine, five received rHuEPO in escalating doses (from 400 to 1,500 U per kg twice daily [BID] per week), alternating with a placebo, in blinded fashion. The second group, consisting of four patients, followed an identical protocol (except starting dose was 1,000 U/Kg, BID per week) and were iron supplemented during treatment. The criterion of response was a transient doubling (as a minimum) of the steady-state F-reticulocyte level. We found that none of the five patients in the first group responded to rHuEPO, and two of them became iron deficient, as judged by a significant decrease in ferritin. Of the second group, four patients responded with F-reticulocyte increases. In three patients, open label administration of rHuEPO confirmed the effect. We observed seven painful episodes during this study, two during the EPO administration and five during the placebo arm. Three patients were phlebotomized because the hemoglobin level increased 1.5 g/dL more than steady-state levels. Of the six patients followed-up by percent dense cell determinations, one exhibited increased levels during periods of the treatment, whereas the other five showed no change. No anti-rHuEPO antibodies were detected. We conclude that rHuEPO can stimulate F-reticulocyte response in some patients with sickle cell anemia, without apparent negative clinical side effects. The state of iron stores may be critical. Whether higher doses of rHuEPO and/or a different regimen might induce sustained F cells and fetal hemoglobin increases remains to be determined.