Objective To establish and verify a risk prediction model for prolonged post-anaesthesia care unit (PACU) stay in patients undergoing radical resection of gastrointestinal neoplasms under general anesthesia, and to provide a valid assessment tool for identifying high-risk patients. Methods A retrospective study was conducted to select 511 surgical patients undergoing radical resection of gastrointestinal neoplasms under general anesthesia from a 3A hospital from June 2020 to December 2021.Patients were divided into a modeling group(361 cases）and a validation group(150 cases).Logistic regression analysis was used to explore the risk factors of prolonged PACU stay. Then a nomogram was drawn, and validated for the model prediction efficacy in the validation group. Results Logistic regression analysis identified such five factors affecting prolonged PACU stay in patients undergoing radical resection of gastrointestinal neoplasms under general anesthesia: ASA grading, hypothermia, hypertension, intraoperative bleeding >200 mL and moderate to severe anemia. Then a risk prediction model of prolonged PACU stay was established according to the five factors. The area under receiver operating characteristic curve (ROC) of the modeling group was 0.805［95%CI (0.752, 0.857)］, and the maximum Youden index was 0.484, versus 0.828［95%CI（0.753，0.914）］ and 0.569 in the validation group. The calibration curve was a straight line with the slope approximately 1. Conclusion The risk prediction model has good predictive efficacy and can provide a reference for early clinical screening of patients at risk of prolonged PACU stay after radical resection of gastrointestinal neoplasms under general anesthesia.