We determined the effect of nocturnal low-flow oxygen (NLFO) on arterial oxygen saturation (SaO2), transcutaneous PCO2 (TcPCO2), and sleep quality in 10 patients with cystic fibrosis (CF) and severe stable chronic obstructive pulmonary disease (COPD). The patients were studied on 2 nights, 1 with oxygen and 1 with air at 2 L/min. The NLFO had no effect upon sleep quality in our patients. The minimal SaO2 occurred during REM sleep and averaged 79.4%. With NLFO, this improved to 92.7%. The average maximal rise in TcPCO2 was 5.6 mmHg on falling asleep while breathing air; this increased a further 5.1 mmHg with NLFO. Two patients also had obstructive sleep apnea. Their SaO2 improved dramatically with NLFO, with no deterioration of ventilation. In 4 patients, ventilation was measured quantitatively. The only consistent changes during air were an increase in abdominal contribution to tidal volume and a drop in minute ventilation from Stage 3-4 to REM sleep of 26%, almost entirely caused by a drop in breathing frequency. The same changes occurred with NLFO. We conclude that NLFO is effective in alleviating the nocturnal hypoxemia of patients with CF with stable COPD and does not cause clinically important hypercapnia.