Objectives: Given that 30-50% of children with Down syndrome have persistent obstructive sleep apnea (OSA) after adenotonsillectomy, we evaluated whether demographic, clinical and polysomnographic factors predicted persistent OSA and OSA severity after adenotonsillectomy. Design: Retrospective study. Setting: Secondary care hospital. Participants: Retrospective review of 32 children with the diagnosis of DS and OSA by polysomnography type 1 who underwent adenotonsillectomy, from January 2010 to December 2018. Main outcome and measure: Non-parametric analysis was used to compare pre and postoperative factors, regression was used to model persistent OSA and OSA severity. Results: Thirty-two children were included (17 male, median age 10.00 ± 8.00 years, median body mass index z-score 0.89 ± 1.25). Overall, adenotonsillectomy resulted in a significant improvement in median obstructive apnea-hypopnea index (oAHI) from 7.5 ± 8.95 to 4.40 ± 4.38 events per hour (p<0.001) and in median OSA-18 score from 85.00 ± 12.00 to 61 ± 37.75 (p<0.001). Persistent OSA was found in 56.25% of the children. Univariate regression suggests that postoperative OSA-18 score predicted persistent OSA after adenotonsillectomy. Preoperative oAHI, preoperative oxygen desaturation rate, pre and postoperative OSA-18 scores correlated with OSA severity after adenotonsillectomy. However, in a multivariate model only the postoperative OSA-18 score was able to predict OSA severity after adenotonsillectomy. Conclusions: Although adenotonsillectomy results in a significant improvement of OSA in children with Down syndrome, more than half of the children had persistent OSA. The postoperative OSA-18 score predicted both persistent OSA and OSA severity after adenotonsillectomy.