Recombinant human erythropoietin (EPO) is an effective and safe therapy for correction of anemia in end-stage renal disease (ESRD). Clinical trials reported to date suggest that EPO improves anemia in over 90% of hemodialysis patients with anemia (Eschbach 1989). Factors which have been identified that appear to inhibit the effectiveness of EPO are infection [Muirhead N et al. 1990], iron deficiency, hyperparathyroidism, aluminium excess [Casati 1991] and persistent GIT bleeding. The development of reticulocytosis in response to EPO in the absence of a rise in hematocrit should alert the clinician to the possibility of either hemolysis or occult blood loss. We present a case in which, despite the development of a reticulocytosis of 5% in response to EPO and the absence of hemolysis, we had difficulty in identifying the presumed source of blood loss.