BACKGROUND Polysomnography findings between matched groups with obstructive sleep apnea (OSA) and OSA plus obesity-hypoventilation syndrome (OHS) before and after continuous positive airway pressure (CPAP), particularly in the extremely severe obese (body mass index [BMI] >or= 50 kg/m2), are unclear. METHODS Prospective study of subjects (BMI >or= 50 kg/m2) undergoing diagnostic polysomnography. Subjects with an apnea-hypopnea index (AHI) >or= 15/h underwent a second polysomnography with CPAP. The effect of 1 night of CPAP on sleep architecture, AHI, arousal indexes, and nocturnal oxygenation was assessed. OHS was defined as those subjects with obesity, PaCo2 > 45 mm Hg, and PaO2 < 70 mm Hg in the absence of lung disease. RESULTS Twenty-three subjects with moderate-to-severe OSA and 23 subjects with moderate-to-severe OSA plus OHS underwent a 1-night trial of CPAP. Both groups were matched for spirometry, BMI, and AHI, but oxygen desaturation was worse in the OSA-plus-OHS group. CPAP significantly improved rapid eye movement (REM) duration (p < 0.005), AHI (p < 0.005), arousal indexes (p < 0.005), and percentage of total sleep time (TST) with oxygen saturation (SpO2) < 90% (p < 0.005) in both groups. In subjects with OSA plus OHS, 43% continued to spend > 20% of TST with SpO2 < 90%, compared to 9% of the OSA group, despite the adequate relief of upper airway obstruction. CONCLUSIONS Extremely severe obese subjects (BMI >or= 50 kg/m2) with moderate-to-severe OSA plus OHS exhibit severe oxygen desaturation but similar severities of AHI, arousal indexes, and sleep architecture abnormalities when compared to matched subjects without OHS. CPAP significantly improves AHI, REM duration, arousal indexes, and nocturnal oxygen desaturation. Some subjects with OHS continued to have nocturnal desaturation despite the control of upper airway obstruction; other mechanisms may contribute. Further long-term studies assessing the comparative role of CPAP and bilevel ventilatory support in such subjects with OHS is warranted.