Chronic Mountain Sickness (CMS) patients have repeatedly been found to hypoventilate. Low saturation in CMS is attributed to hypoventilation. Although this observation seems logical, a further understanding of the exact mechanism of hypoxia is mandatory. An exercise study using the Bruce Protocol in CMS (n = 13) compared to normals N (n = 17), measuring ventilation (VE), pulse (P), and saturation by pulse oximetry (SaO(2)) was performed. Ventilation at rest while standing, prior to exercise in a treadmill was indeed lower in CMS (8.37 l/min compared with 9.54 l/min in N). However, during exercise, stage one through four, ventilation and cardiac frequency both remained higher than in N. In spite of this, SaO(2) gradually decreased. Although CMS subjects increased ventilation and heart rate more than N, saturation was not sustained, suggesting respiratory insufficiency. The degree of veno-arterial shunting of blood is obviously higher in the CMS patients both at rest and during exercise as judged from the SaO(2) values. The higher shunt fraction is due probably to a larger degree of trapped air in the lungs with uneven ventilation of the CMS patients. One can infer that hypoventilation at rest is an energy saving mechanism of the pneumo-dynamic and hemo-dynamic pumps. Increased ventilation would achieve an unnecessary high SaO(2) at rest (low metabolism). This is particularly true during sleep.