Objective: To construct a nomogram prediction model for the risk of stroke in patients with hip fractures during the perioperative period. Methods: The clinical data of patients with hip fracture surgery admitted to a Class III Grade A hospital from January 2019 to December 2023 were selected (n = 895). According to the presence or absence of perioperative stroke, the patients were divided into the case group (n = 53) and the control group (n = 842). Univariate and multivariate analyses were used to analyze the risk factors of stroke in patients, and a prediction model was constructed. A nomogram was drawn to visualize the logistic regression prediction model, and the ROC curve and decision curve were used to analyze the prediction efficiency of the logistic regression model. Results: Age ≥60 years, BMI ≥24 kg/m², history of hypertension, history of diabetes, history of stroke, general anesthesia, abnormal intraoperative blood pressure, and operation time were included in the final prediction model. The AUC of the validation set was 0.955 (95% CI: 0.923–0.986). The calibration curve showed that the predicted probability fitted the actual incidence well (Hosmer-Lemeshow test P = 0.005), and DCA confirmed that the model had a net clinical benefit. Conclusion: The risk prediction model constructed in this study can effectively identify patients with a high risk of perioperative stroke after hip fracture surgery, and has good prediction efficiency and clinical practicability, which is convenient for early and accurate prediction of perioperative stroke risk.