Objective To establish the risk prediction scoring model of pulmonary infection(PI) in cerebral hemo-rrhage patients undergoing surgery, identify the high-risk population of PI, and provide basis for health care wor-kers(HCWs) to take early effective prevention and control measures. Methods Clinical data of cerebral hemorrhage patients undergoing surgery in a hospital in Shandong Province in 2016-2018 were collected prospectively, patients were randomly divided into modeling group and validation group based on a ratio of 7∶3, logistic regression model was established based on data of modeling group, factors were assigned according to β value, risk prediction scoring model of PI was established. The established risk scoring model was used to score the risk of patients in modeling group and validation group, and the predictive ability of the model was judged by receiver operating characteristic (ROC) curve and Hosmer-Lemeshow(H-L) test. Results Logistic regression analysis showed that endotracheal intubation, indwelling gastric tube, venous thrombosis and ASA were independent risk factors for PI in cerebral hemorrhage patients undergoing surgery, and corresponding scores in the risk scoring model were 5, 6, 4, and 2 points (per stratification) respectively. AUC under ROC curve of the risk scoring model in the modeling group was 0.864 (95%CI: 0.825-0.904, P < 0.001), H-L test was P=0.404; AUC under ROC curve in validation group was 0.861(95%CI: 0.800-0.921, P < 0.001), H-L test was P=0.452. Conclusion The established risk prediction scoring model can better predict the risk of PI in cerebral hemorrhage patients undergoing surgery, which can early identify high-risk population and optimize the prognosis of patients.