Objective:To establish and validate the prediction model for postoperative sleep disturbance (PSD) in patients undergoing non-cardiac surgery.Methods:A total of 454 patients of both sexes, aged≥18 yr, of American Society of Anesthesiologists physical statusⅠ-Ⅲ, underwent non-cardiac surgery under general anesthesia from November 2019 to September 2020 were selected.The perioperative data were collected.The patients were divided into training set and validation set with a ratio of 7∶3 by using a simple random sampling method.The characteristic variables of PSD were selected using LASSO regression analysis and the independent risk factors were identified using multivariate logistic regression analysis in training set.Akaike′s information criterion was used to evaluate the quality of fit of the model.The nomogram of PSD in non-cardiac surgery patients was constructed based on the identified factors.The discrimination of the model was evaluated using receiver operating characteristic (ROC) curve, and the agreement of the model was evaluated using Hosmer-Lemeshow goodness-of-fit test and Brier score.Results:Seven risk factors (gender, preoperative anxiety, satisfaction with the ward environment, anesthesia time, the intraoperative consumption of midazolam and sufentanil and numerical rating scale (NRS) score at 3 days after operation) and two related factors (preoperative NRS score and general anesthesia combined with nerve block) were used to establish and verify the PSD nomogram.The area under the ROC curve was 0.805 (95% confidence interval [CI] 0.721-0.848) in training set.The area under the ROC curve was 0.773 (95% CI 0.684-0.876) in validation set.In training and validation sets, the calibration curves were tested by Hosmer-Lemeshow good of fit test, the P values were 0.590 and 0.950, respectively, and the Brier scores were 0.154 and 0.156, respectively.The nomogram predicated that the sensitivity (95% CI) and specificity (95%CI) were 81.83% (60.32%-95.14%) and 78.15% (71.83%-83.25%), respectively, in training set, and the sensitivity (95% CI) and specificity (95%CI) were 77.86% (39.84%-97.25%) and 78.15% 77.86% (68.74%-85.48%), respectively, in validation set.The optimal cut-off value of nomogram score was 113. Conclusion:In this study, the nomogram prediction model for PSD in patients undergoing non-cardiac surgery has been successfully established, which can visually and individually predict the risk of PSD.