Objective: Depression is a common problem in patients with obstructive sleep apnea. The objective of this study was to examine whether depression is independently associated with lower self-reported sleep quality in patients with obstructive sleep apnea (OSA), after controlling for polysomnographic measures of sleep. Methods: The sample comprised 135 patients who had been referred to a university teaching hospital's multidisciplinary sleep medicine center for polysomnographic evaluation of OSA. The median age of the subjects was 45 (mean age, 46 years) 55% were female, 69% were white, 31% were black, and their mean body mass index was 37.9 +/- 11.2 kg/m(2). Self-reported sleep quality during the past 2 weeks was assessed by the insomnia severity index. Polygraphic measures of sleep quality included the respiratory disturbance index, sleep onset latency, arousals for no apparent reason, sleep efficiency, and periodic leg movements associated with arousal. Depressive symptoms were assessed by the Beck Depression Inventory. Results: None of the polygraphic measures of sleep quality was related to self-reported sleep quality or depression. Oxygen desaturation was correlated with self-reported sleep quality (r = 0.21, p = .02). Depression correlated with self-reported sleep quality (r = 0.55, p <.0001). In a multiple regression analysis, depression remained a significant predictor of self-reported sleep quality after controlling for all of the polysomnographic measures of sleep quality (F = 9.65, partial r(2) = 0.28 p = .0001). Conclusion: Depression is a better predictor of self-reported sleep quality than are polysomnographic measures of sleep in patients with OSA.