The “living high–training low” model (Hi–Lo) may improve aerobic performance in athletes, and the main mechanism of this improvement is thought to be augmented erythropoiesis. A positive effect of Hi–Lo has been demonstrated previously by using altitudes of 2,000–3,000 m. Since the rate of erythropoiesis is altitude-dependent, we tested whether a higher altitude (3,500 m) during Hi–Lo increases erythropoiesis and maximal aerobic performance. Nordic skiers trained for 18 days at 1,200 m, while sleeping at 1,200 m in ambient air (control group, n = 5) or in hypoxic rooms (Hi–Lo, n = 6; 3 × 6 days at simulated altitudes of 2,500, 3,000 and finally 3,500 m, 11 h day⁻¹). Measurements were done before, during (blood samples only) and 2 weeks after the intervention (POST). Maximal aerobic performance was examined from [(V)\dot]O₂max\dot{V}\hbox{O}_{2}\hbox{max} and time to exhaustion (T _exh) at v[(V)\dot]O₂maxv\dot{V}\hbox{O}_{2}\hbox{max} (minimum speed associated with [(V)\dot]O₂max\dot{V}\hbox{O}_{2}\hbox{max}), respectively. Erythropoietin and soluble transferrin receptor responses were higher during Hi–Lo, whereas reticulocytes did not change. In POST (vs. before): hematological parameters were similar to basal levels, as well as red blood cell volume, being 2.68 ± 0.83 l (vs. 2.64±0.54 l) in Hi–Lo and 2.62±0.57 l (vs. 2.87 ± 0.59 l) in controls. At that time, neither [(V)\dot]O₂max\dot{V}\hbox{O}_{2}\hbox{max} nor T _exh were improved by Hi–Lo, [(V)\dot]O₂max\dot{V}\hbox{O}_{2}\hbox{max} being non-significantly decreased by 2.0% (controls) and 3.7% (Hi–Lo). The present results suggest that increasing the altitude up to 3,500 m during Hi–Lo stimulates erythropoiesis but does not confer any advantage for maximal O₂ transport.