Administration of erythropoietin (EPO) has been linked to cerebrovascular events. EPO reduces vascular conductance, possibly because of the increase in hematocrit. Whether EPO in itself affects the vasculature remains unknown; here it was evaluated in healthy males by determining systemic and cerebrovascular variables following acute (30,000 IU/d for 3 d; n=8) and chronic (5000 IU/week for 13 wk; n=8) administration of EPO, while the responsiveness of the vasculature was challenged during cycling exercise, with and without hypoxia. Prolonged administration of EPO increased hematocrit from 42.5 ± 3.7 to 47.6 ± 4.1% (P<0.01), whereas hematocrit was unaffected following acute EPO administration. Yet, the two EPO regimes increased arterial pressure similarly (by 8±4 and 7±3 mmHg, respectively; P=0.01) through reduced vascular conductance (by 7±3 and 5±2%; P<0.05). Also, both EPO regimes widened the arterial-to-jugular O(2) differences at rest as well as during normoxic and hypoxic exercise (P<0.01), which indicated reduced cerebral blood flow despite preserved dynamic cerebral autoregulation, and an increase in middle cerebral artery mean blood flow velocity (P<0.05), therefore, reflected vasoconstriction. Thus, administration of EPO to healthy humans lowers systemic and cerebral conductance independent of its effect on hematocrit.