The objective of this study was to quantitate the effects of O2 supplementation by nasal cannula (NC) and Venturi mask (VM) on PaO2 in patients with chronic obstructive pulmonary disease (COPD) during acute hypobaric exposure, simulating a commercial jet aircraft cabin. We conducted a crossover intervention trial in which subjects served as their own controls in an ambulatory outpatient pulmonary disease service of a tertiary care military medical center and a hypobaric research facility. The subjects were a volunteer sample of 18 men with stable severe COPD, not requiring long-term O2 therapy, and uncomplicated by hypercapnea or cardiac disease. Mean age was 68 years, and mean FEV1 was 0.97 L (31.3 percent predicted). We exposed patients to conditions equivalent to 8,000 feet in a hypobaric chamber. Radial artery catheters provided blood samples at ground level and 8,000 feet. O2 was sequentially administered at 8,000 feet by NC at 4 L/min and 24 percent or 28 percent VM. We describe changes in blood gas data from baseline values and between interventions. O2 at 4 L/min NC flow at 8,000 feet caused PaO2 to increase from 47.4 +/- 6.3 mm Hg to 82.3 +/- 14 mm Hg (n = 18), an increase of 34.9 +/- 14.8 mm Hg. Supplementation of O2 by 24 percent VM caused PaO2 at 8,000 feet to increase by 12.7 +/- 3.8 mm Hg. Twenty-eight percent VM caused PaO2 at 8,000 feet to increase by 19.7 +/- 8.2 mm Hg. Changes in PaO2 with 4 L/min NC were greater than those with either VM. The increase with 28 percent VM was greater than that caused by 24 percent VM (p less than 0.05). Compared with ground level, 4 L/min NC increased mean PaO2 by 9.9 +/- 12.6 mm Hg; 24 percent and 28 percent VM did not cause mean PaO2 to increase above ground level values. We describe a range of capability of familiar O2 therapy devices to increase PaO2 to levels that will maintain tissue oxygenation of patients during acute altitude exposure.