The hypothesis that supplemental oxygen could improve the quality of sleep was tested in 23 consecutive patients (14 male, nine female; age 42-74 years) with chronic obstructive lung disease (mean (SD) FEV1 0.81 (0.32) litre, FEV1/FVC 37% (12%). Patients breathed compressed air or supplemental oxygen via nasal cannulas on consecutive nights in a randomised, double blind, crossover trial. Quality of sleep was assessed by questionnaire and by electroencephalographic sleep staging. The study had a power of 80% to detect, at the 0.05 level, a 20% improvement in total sleep time. Seventeen patients slept for two nights in the laboratory. Oxygenation during sleep was improved by oxygen administration, but there was no improvement in quality of sleep. There was an acclimatisation effect with better sleep on the second night. Six patients spent an additional acclimatisation night in the laboratory as well as the two study nights. There was no difference in sleep quality between the second and third nights or between the compressed air and the oxygen nights in these patients. Subgroups of patients with an arterial carbon dioxide tension of over 43 mm Hg (5.7 kPa) (n = 12) and arterial oxygen saturation of less than 90% (n = 11) while awake did not show any improvement in quality of sleep on the oxygen night. It is concluded that supplemental oxygen improves nocturnal oxygenation but does not immediately improve the quality of sleep in the laboratory in patients with chronic obstructive lung disease.