We have compared the efficacy of oral to intravenous iron for the chronic maintenance of iron stores in hemodialysis patients. Fifty-two hemodialysis patients with initial serum ferritin greater than 100 ng/mL and transferrin saturation greater than 15% were randomly assigned to one of two groups: those receiving oral iron therapy (n = 32) and those receiving intravenous iron dextran (100 mg twice weekly) (n = 20). At study completion (4 months), the mean hematocrit was significantly higher in the intravenous group than in the oral iron group (34.4% +/- 0.7% v 31.8% +/- 0.4%, respectively; P < 0.05), the final mean recombinant human erythropoietin dose was 46% lower in the intravenous iron group than in the oral group (4,050 +/- 634 U/treatment v 7,563 +/- 378 U/treatment; P < 0.05), and the mean serum ferritin was significantly higher in the intravenous group than in the oral iron group (753.9 +/- 30.2 ng/mL v 157.3 +/- 15.4 ng/mL, respectively; P < 0.05). We have found that administering iron intravenously instead of orally for chronic maintenance iron supplementation in hemodialysis patients resulted in improved erythropoiesis. We hypothesize that most hemodialysis patients have inadequate iron stores for optimal erythropoiesis when currently recommended levels of ferritin and transferrin saturation are used to guide therapy, and that the chronic use of intravenous iron could reduce recombinant human erythropoietin requirements by maximizing iron stores. The improvement in erythropoiesis was accompanied, however, by an increase in iron indices to levels that could be indicative of tissue iron overload. Future studies must be performed to determine whether lower doses of intravenous iron dextran would improve erythropoiesis without causing potential organ iron overload.