The "living high-training low" (LHTL) model is frequently used to enhance aerobic performance. However, the clinical tolerance and acclimatization process to this intermittent exposure needs to be examined. Forty one athletes from three federations (cross-country skiers, n=11; swimmers, n=18; runners, n=12) separately performed a 13 to 18-day training at the altitude of 1,200 m, by sleeping either at 1,200 m (CON) or in hypoxic rooms (HYP), with an O2 fraction corresponding to 2,500 m (5 nights for swimmers and 6 for skiers and runners), 3,000 m (6 nights for skiers, 8 for swimmers and 12 for runners) and 3,500 m (6 nights for skiers). Measurements performed before, 1 or 15 days after training were ventilatory response (HVRe) and desaturation (deltaSaO2e) during hypoxic exercise, an evaluation of cardiac function by echocardiography, and leukocyte count. Lake Louise AMS score and arterial O2 saturation during sleep were measured daily for HYP. Subjects did not develop symptoms of AMS. Mean nocturnal SaO2 decreased with altitude down to 90% at 3,500 m and increased with acclimatization (except at 3,500 m). Leukocyte count was not affected except at 3,500 m. The heart function was not affected by LHTL. Signs of ventilatory acclimatization were present immediately after training (increased HVRe and decreased deltaSaO2e) and had disappeared 15 days later. In conclusion, LHTL was well tolerated and compatible with aerobic training. Comparison of the three patterns of training suggests that a LHTL session should not exceed 3,000 m, for at least 18 days, with a minimum of 12 h day(-1) of exposure.