Objective: To analyze the clinical and prognostic characteristics of rapid eye movement sleep related obstructive sleep apnea (REM-OSA) in children. Methods: A retrospective analysis was performed on the clinical data of 62 children aged from 2 to 14 years who were admitted to Beijing Children's Hospital, Capital Medical University from December 2017 to April 2021, diagnosed with moderate to severe OSA by polysomnography monitoring (PSG), underwent adenoid tonsillectomy, and completed follow-up 6 months after surgery. There were 45 males (72.6%) and 17 females (27.4%). The age range was 2.0-12.3 years. All children completed the clinical data collection, PSG, OSA-18 quality of life questionnaire and Children's Sleep questionnaire-sleep related breathing disorder subscale at baseline. PSG and OSA-18 quality of life questionnaire were reexamined at 6 months after surgery. Children were divided into REM-OSA group (33 cases) and non-REM-OSA group (29 cases) according to whether the obstructive apnea-hypopnea index (OAHI) during rapid eye movement sleep and OAHI during non-rapid eye movement sleep ratio was≥2. Baseline PSG parameters and scale scores, 6-month postoperative cure rate and OSA-18 quality of life questionnaire scores of the 2 groups were compared, and statistical analysis was performed using SPSS 23.0 software. Results: There were no significant differences in age, sex, body mass index, neck circumference/height ratio, overweight or obesity, history of disease, tonsil and adenoid size between the two groups (all P>0.05). Compared with non-REM-OSA group, REM-OSA group had higher oxygen desaturation index and proportion of SpO2<90% of total sleep time (Z=-2.723, P=0.006;Z=-3.414; P=0.001 respectively), and lower SpO2 nadir (Z=-3.957, P<0.001). The proportion of obstructive apnea in total respiratory events (related to anatomical factors) in REM-OSA group was higher than that in non-REM-OSA group (t=2.840, P=0.006). However, the proportion of central apnea in total respiratory events and arousal index (related to functional factors) in REM-OSA group was lower than that in non-REM-OSA group (t=-2.597, P=0.012;Z=-2.956, P=0.003), and there were no significant differences in other PSG parameters between the two groups (all P>0.05). There was an interaction effect between the two groups in the change trend of OSA cure rate at 6 months after surgery under different baseline OAHI (χ2=4.282, P=0.039). Conclusions: The weight of anatomic factors and functional factors in the etiology of children with REM-OSA and non-REM OSA was different, and the postoperative OSA cure rate of children with different baseline OAHI changed in reverse trend.