Recombinant human erythropoietin (EPO) increases exercise capacity by stimulating erythropoiesis and subsequently enhancing oxygen delivery to the working muscles. In a large dose, EPO cross the blood brain barrier and may reduce central fatigue and improve cognition. In turn, this would augment exercise capacity independent of erythropoiesis. To test this hypothesis, 15 healthy young males (18-34 yo., 74±7 kg) received either 3 days of high dose (30,000 IU day-1, N=7) double-blinded placebo controlled or 3 months of low dose (5,000 IU week-1, N=8) counter-balanced open but controlled administration of EPO. Administration of high, but not low, doses of EPO increased cerebrospinal fluid EPO concentration increased from 0.6±0.2 to 13.9±4.0 mIU ml-1 (P<0.05). High dose EPO administration increased tidal volume (from 2.8±0.3 to 3.0±0.3 l) and respiration frequency (34±5 to 38±6 breath min-1) during exercise in hypoxia. The increase in pulmonary ventilation was accompanied by an increase in arterial oxygen saturation from 58±5 to 62±4% (P<0.01). No changes in pulmonary ventilation or arterial oxygen saturation were observed with low dose, prolonged EPO administration. No significant changes were observed in cerebral glucose metabolism. In contrast, cerebral arterio-venous lactate difference decreased ~0.3 mmol l-1 during exercise with high dose, short term EPO administration (P<0.05), while no changes were observed with low dose, prolonged EPO administration. Ratings of perceived exertion (RPE) was lowest with low intensity exercise and increased with addition of hypoxia or when exercise was maximal (P<0.05). RPE decreased with prolonged low dose EPO administration (P<0.05), but increased with high dose, short term EPO administration (P<0.05). Voluntary activation of the elbow flexors decreased during exercise in both groups (from 98.1±3.9% to 96.2±1.6% , P<0.001), however, EPO did not have a significant effect. There were 33 evaluations of cognitive function and strategy and of those variables, no one was found to change significantly with EPO treatment in either group. Exercise performance increased ~15% with 3 months of prolonged EPO administration from 511±129 to 600±175 kJ (P<0.01). In contrast, exercise performance was unchanged with high dose, short term EPO administration (423±90 vs. 436±55 kJ). In conclusion, we confirmed that 3 month’s administration of EPO increases exercise capacity, but the improvement could not be accounted for by other mechanisms than enhanced oxygen delivery. Also, EPO does not attenuate central fatigue or changes cognitive performance strategy suggesting that EPO enhances exercise capacity exclusively by increased oxygen delivery to the working muscles.