Background: Positive airway pressure (PAP) therapy is the standard treatment for obstructive sleep apnea syndrome (OSAS). Objectives: The aim of the current study was to determine operational long-term adherence to PAP and its predictors. Methods: In a retrospective single-center observational cohort study, we analyzed all patients referred to our center with suspected OSAS between November 2001 and November 2011. Baseline results and last follow-up data of each patient were analyzed. Kaplan-Meier estimates of adherence and Cox proportional hazard regression for age, gender, Epworth sleepiness scale (ESS) scores, body mass index, apnea-hypopnea index (AHI) and oxygen desaturation index (ODI) were performed. Evolution of adherence was analyzed in yearly cohorts comparing the proportion of patients discontinuing PAP within 6 and 12 months. Results: Of 4,638 referrals, 2,187 confirmed OSAS patients started PAP, 297 (14%) were referred out to other centers to follow-up, 42 (2%) died, and 92 (5%) no longer needed PAP. Of 1,756 patients, the median follow-up was 36 months [95% confidence interval (CI) 33.6-37.8], and adherence at 1, 5 and 10 years was 74 (CI 71-75; n = 1,028), 55 (CI 53-58; n = 281) and 51% (CI 48-55; n = 10), respectively. Adherence is associated with ESS score [hazard ratio (HR) 0.60; CI 0.47-0.78], ODI (HR 0.50; CI 0.32-0.77) and AHI (HR 0.56; CI 0.37-0.85). In yearly cohorts according to inclusion date, the absconder rate at 6 and 12 months was 20 (CI 18-22) and 27% (CI 25-30) for the first 8 years and improved to 10 (CI 7-15) and 14% (CI 10-19) for the last 2 years, respectively. Conclusions: Long-term adherence to PAP in OSAS is associated with baseline measures of disease severity. After 2009, an improvement in the adherence rate was observed.