Introduction The effectiveness of continuous positive airway pressure (PAP) therapy is primarily measured by the reduction of the apnea-hypopnea index (AHI). A proxy measure of the sleep study-derived AHI can be obtained from the PAP device during the portion of the sleep period during which it is worn. Sleep quality and daytime functioning are considered two of the main outcomes desired by patients. Surprisingly few studies have examined the effectiveness of PAP therapy on sleep quality. Methods OSA participants (n = 695) from a combination of larger trials that examined a PAP adherence intervention were included. Participants were provided with PAP instruction and followed at 2 months. The Pittsburgh Sleep Quality Index (PSQI) was used as the primary measure of sleep quality. Results The PSQI total score was significantly correlated with PAP adherence at the 2-month time point, such that lower sleep quality was associated with lower PAP use. This finding held for the sleep disturbance subscale of the PSQI. The total PSQI score at baseline was 12.8±3.4 and at 2-month follow-up was 9.7±3.6, which is over the threshold of 5 for the PSQI total score and indicates poor sleep quality. Over 52% of those using PAP therapy at the 2-month time point reported significantly disturbed sleep, with the top three causes: 1) Wake up in the middle of the night or early morning (59%); 2) Have to get up to use the bathroom (56%); and 3) Have pain (33%). Conclusion This study shows that PAP therapy does not improve sleep quality to an acceptable degree. Over 50% patients using PAP therapy still experienced disturbed sleep. Whether the disturbed sleep is directly attributable to the PAP device itself or to disturbed sleep secondary to uncontrolled OSA when PAP is not worn requires further investigation. Clinical practice needs to focus on patient outcomes and not a single proxy measure of device effectiveness. Support (If Any) This project was supported in part by Department of Veteran Affairs and VA San Diego Healthcare System Research Service.