We investigated nocturnal oxygen desaturation (NOD) in 36 patients with stable chronic respiratory disease who were receiving home oxygen therapy (HOT). Study data included medical history, chest roentgenograms, measurement of daytime arterial blood gases while awake, and spirometry. Each subject underwent full overnight oximetry monitoring. Three patients were excluded from further investigation because of periodic desaturation suggestive of sleep apnea. The remaining 33 subjects were divided into two groups: 21 patients with sequelae of pulmonary tuberculosis (TB-sequela) and 12 patients with chronic obstructive pulmonary disease (COPD). The COPD group was divided into two subgroups according to the Burrows classification (Am Rev Resp Dis. 90: 14-27, 1964): 5 patients with type A (Type A) and 7 patients with type B (Type B) COPD. The percentages of total sleep time with SaO₂≤ 85% (DST85) and SaO_<2≤ 90% (DST90) were calculated for each subject. NOD was defined as DST85≥ 1%. Arterial oxygen partial pressure (PaO₂) while awake was≥ 60 Torr in all subjects. No difference was observed in mean awake PaO₂values between the TB-sequela and COPD groups. NOD was detected in 8 TB-sequela patients but in none of the COPD patients. Mean DST85 and DST90 values were significantly (p<0.05) higher for the TB-sequela group than for the COPD group. Of 15 TB-sequela patients who were able to complete spirometry tests, 6 had NOD. All 6 of these patients had hypercapnia while awake (PaCO₂≥ 50 Torr) and reduced vital capacity (≤ 50% predicted). No difference was observed in mean DST90 or DST85 values between the TypeA and TypeB COPD subgroups. We conclude that NOD is common in patients with chronic stable respiratory disease treated with HOT despite daytime euoxia. TB-sequela patients with hypercapnia and restrictive ventilatory impairment are at high risk for NOD.