Introduction: Coaches and athletes use hypoxic training (Live low, Train high), convinced of performance benefits over similar training at sea level. The purpose of this investigation was to document a performance benefit with a commonly used hypoxic training protocol in a blinded randomized cross-over design. Further, this study investigated several possible mechanisms for any improvements observed. We hypothesized that 3 weeks of hypoxic interval trainin would increase sea level performance and VO2 max compared to training at the same relative workload at sea level. Methods: 8 junior elite cyclists who were sea level residents (7 men, 1 woman, age 17 ± 1 years, weight 70.6 ± 7.8 kg, height 180 ± 5 cm) were recruited from the Norwegian high school for top athletes. Each subject underwent 2, three-week randomly assigned, training periods in our ônitrogen roomsö, one period in normoxia and the other in normobaric hypoxia (15% O2). A 5-week ôwashout periodö occurred between the training regimes. Each training period consisted of 9, 2-hour interval workouts, 3 per week at the subjects' individual La threshold. Heart rate, lactate, SpO2 and work (Kjoules) were recorded. Before and after each 3-week training period, the subjects performed a 30-minute time trial on their racing bike, a series of 5 submaximal workloads at increasing intensity and a VO2 max text on a stationary bike. Measurements done during the tests included VO2, ventilation, RQ, cardiac output, heart rate, lactate, SpO2, and arterial blood gases. Blood volume plasma volume and total red cell colume were measured at each time point (Evans blue dye). Results: During training there was a significant difference in absolute workload and oxygen content in the blood between the normoxic and the hypoxic group, while there were no differences in mean heart rate or lactate accumulated during training. Workload ± SD (watt): Normoxia: 245 ± 37, Hypoxia: 206 ± 32*. SpO2 ± SD (%): Normoxia: 96 ± 1, Hypoxia: 84 ± 3*. HR ± SD (bt/min: Normoxia: 163 ± 8, Hypoxia: 160 ± 7. Lactate ± SD (Mmol: Normoxia: 2.9 ± 0.2, Hypoxia: 3.2 ± 0.3. No significant differences were found in hematologic parameters at any time point (period 1 vs 2 or hypoxic vs normoxic). No significant changes were noted in maximal: power, VO2, VE (BTPS), heart rate, lactate, SpO2, RQ or time trial performance at any time point (period 1 vs 2 or hypoxic vs normoxic). Conclusions: We therefore conclude that 3 weeks of interval training at a simulated altitude of 3000 meter above sea level does not give an additional performance benefit at sea level compared to the same relative training a sea level. The hypoxic exposure is too brief to provide for any hematologic changes.