Exercising in high altitude is impeded during the first days of exposure to altitude hypoxia by the symptoms of Acute Mountain Sickness (AMS). Susceptibility to AMS is independent of endurance training and determined by the sensitivity of carotid chemoreceptors to hypoxemia and induced hyperventilation and tachycardia. Cardiac response to hypoxia is not as well known as ventilatory response, especially at exercise. A group of 138 male alpinists has been explored before their departure to a high altitude expedition using an hypoxic gas mixture (equivalent altitude = 4.800 m), at rest and at exercise (5 minutes at 50 p. 100 maximal O2 consumption). Cardiac response to hypoxia was assessed by the ratio DFc/DSaO2: variation in heart rate (hypoxia - normoxia)/variation in arterial O2 saturation, at rest (r) and during exercise (e). After the expedition, subjects were classified in AMS+ or AMS- group if they suffered or not from severe AMS, and also following their climbing skill. Cardiac response to hypoxia at rest is lower in AMS+ subjects (DFc/DSaO2 = - 0.86 +/- 0.40 nn - 1.% - 1) than in AMS- subjects (-1.12 +/- 0.69, p less than 0.05), but is not linked to the climbing skill. Similar differences were observed at exercise: DFc/DSaO2e = - 0.88 +/- 0.32 (AMS+) and - 1.05 +/- 0.50 (AMS-) (p less than 0.05). Associated with the respiratory response to hypoxia, the cardiac response allows the detection of AMS high-risk subjects and may be used in an much less than aptitude to altitude much greater than test.