Experiments were done to investigate the effects of increased inspired oxygen on the arousal response from sleep to upper airway obstruction in 10 newborn lambs. Each lamb was anesthetized and instrumented for recordings of electrocorticogram, electrooculogram, nuchal and diaphragm electromyograms, and measurements of systemic arterial blood pressure and oxygen saturation (fiberoptic catheter oximeter). A tracheotomy was performed and a fenestrated tracheostomy tube was placed in the trachea. A 5F balloon tipped catheter was inserted into the tube so that airflow could be obstructed by inflating the balloon. No sooner than 3 days after surgery, measurements were made during a control period and during an experimental period of upper airway obstruction; the inspired oxygen fraction was alternated hourly between 0.21 and 0.60. A total of 57 epochs of quiet sleep and 58 epochs of active sleep was obtained in eight lambs. Arousal was significantly delayed (p < 0.005) during active sleep (21 ± 6 s; mean ± 1 SD) compared to quiet sleep (7 ± 2 s) in room air. Increased inspired oxygen significantly delayed arousal (p < 0.05) during active sleep (47 ± 25 s), but had little effect on arousal in quiet sleep (10 ± 4 s). These results provide evidence that arousal from active sleep following upper airway obstruction in lambs is primarily initiated by a decrease in arterial oxygen. However, arousal from quiet sleep following upper airway obstruction in lambs appears to be initiated by other stimuli.