Abstract Introduction This study assessed the frequency and duration of arousal responses associated with posture shifts during sleep and the time taken to resume sleep following posture shifts with versus without a supine-avoidance alarm. Method Sleep and posture shift outcomes during a baseline in-home sleep study (Embletta MPR plus an inactive supine-avoidance alarm device) and subsequent sleep study after 6-8 weeks of supine-avoidance treatment were examined in 66 patients with supine-predominant OSA. A technician blinded to supine-treatment data scored each sleep study, including arousal responses (3-15 and ≥15-sec EEG changes). All posture shifts lasting >5 sec were identified from sleep study and supine-alarm device posture recordings. All sustained (≥30 sec) posture shifts commencing from established sleep during baseline and supine-avoidance nights were assessed for arousal responses (within -30 to +3 sec of each posture shift) and the time taken to resume sleep following each posture shift. Results The study population was 62.1% males with a mean±SD age 52.8±11.9 years, BMI 31.9.4±7.8 kg/m², AHI 18.0±9.9 /hr (supine AHI=40.2±38.8 /h). From a total of 418 (baseline) vs 283 (supine-avoidance) sustained posture shifts (median [IQR] 6[3-10] vs 6[3-9] per patient), the majority [342/418 vs 244/283; 91[75-100] vs 100[84-100]%) were associated with an arousal, followed by 0.5 [0.0 to 1.0] min before sleep resumed, with no difference with versus without the supine-alarm. Conclusion Posture shifts from sleep are mostly associated with brief arousal prior to any supine-avoidance alarm, after which a supine avoidance alarm effectively discourages supine positioning with minimal disturbance to the resumption of sleep.