Increasing hypoxia with altitude ascent is a potentially serious problem for patients with hypoxemic chronic airway obstruction (CAO) at sea level. We developed a hypoxia-altitude simulation test (HAST) to assess acute cardiopulmonary responses to the inhalation of hypoxic gas mixtures (equivalent to the inspired oxygen tension (PO2) present at 5,000, 8,000, and 10,000 feet above sea level) alone and in combination with supplemental oxygen (O2). Twenty-two subjects with stable normocapnic CAO were studied at sea level with a computer-based system that measured on-line, breath-by-breath resting ventilatory and gas exchange variables. Subjects breathed 20.9% (baseline), 17.1, 15.1, 13.9, and 20.9% (recovery) O2, and measurements were obtained once a "steady state" was reached at each level. Steady-state arterial PO2 (PaO2) and O2 saturation, alveolar PO2, and alveolar-to-arterial PO2 gradient decreased markedly during successive hypoxic levels, whereas arterial carbon dioxide tensions decreased only modestly. Minute ventilation and heart rate during 13.9% O2 increased only 12 and 10% above baseline. Ten subjects had asymptomatic cardiac arrhythmias during the HAST. Supplemental O2 significantly improved nearly all physiologic indexes. Sea level PaO2 best predicted acute, resting altitude PaO2. Sea level PaO2 values of 68 and 72 mmHg successfully classified more than 90% of the subjects with a PaO2 greater than 55 mmHg at 5,000 feet and a PaO2 greater than 55 mmHg at 8,000 feet, respectively. A regression equation and nomogram were derived to estimate PaO2 at altitudes between 5,000 to 10,000 feet in patients with normocapnic CAO.(ABSTRACT TRUNCATED AT 250 WORDS)