OBJECTIVE To test the ability to elicit a hemoglobin (Hb) response in patients on chronic hemodialysis, we prospectively compared two regimens of iron dextran administration, 100 mg once weekly (QW) or 100 mg once every dialysis (QD), both given for 10 doses. METHODS Twenty-three consecutive patients on chronic hemodialysis received iron dextran intravenously if they had absolute or functional iron deficiency. There was no difference in the Hb response between regimens. RESULTS Both groups had a significant increase in Hb from 10.5+/-1.5 g/dl at baseline, to 11.1+/-1.7 g/dl at 1 month, 1.4+/-2.1 g/dl at 2 months and 11.6+/-1.9 g/dl at 3 months. The increment in Hb at 1 month was similar (QD 0.62+/-1.245 g/dl vs. QW 0.64+/-1.464 g/dl) between the two groups despite a large difference in the amount of iron received. Serum ferritin, transferrin saturations or epoetin dose did not change significantly. At the end of 3 months 12 patients did not need further iron therapy as judged by the serological markers of iron stores. Of these 12 patients, 3 had serum ferritins of > 1,000 ng/ml. Weekly dosing of iron was associated with more medication errors than dosing every dialysis. Baseline iron stores could not predict the responsiveness to intravenous iron therapy as judged by an increase in Hb concentration at 1 month or at 3 months. CONCLUSIONS This study confirms the efficacy of 1,000 mg of intravenous iron administered over a 3-month period in patients with functional iron deficiency. It underscores the importance of careful monitoring of iron stores and highlights the need for developing better parameters of functional iron stores in hemodialysis patients.