Objective To analyze the risk factors of perioperative blood transfusion in elderly femoral intertrochanteric fractures, and establish a logistic regression prediction model to guide the clinical prediction of blood transfusion risk. Methods A retrospective case-control study was used to analyze the clinical data of 174 patients (age 60 years or older) treated with intertrochanteric fractures treated from August 2015 to January 2020, including 54 males and 120 females, age 80.5 ± 8.4 years old. The patients were divided into transfusion group and non-transfusion group, and the independent risk factors of blood transfusion were obtained by univariate and multivariate analysis. The perioperative complications were compared between the two groups. A logistic regression prediction model was established, and finally the Hosmer and Lemeshow test was used to test the goodness of fit of the prediction model, and the area under the ROC curve was used to evaluate the effectiveness of the prediction model. Results The univariate analysis showed that the two groups were different in age, fracture type, American Society of Anesthesiologists (ASA) classification, length of operation, intraoperative blood loss, whether to expand the medulla, hemoglobin (Hb), hematocrit (HCT), The differences in creatinine (Cr) and other aspects were statistically significant (P <0.05). The results of multivariate analysis showed that the elderly (OR = 1.066, CI = 1.011 ～ 1.124) and more intraoperative bleeding (OR = 1.007, CI = 1.003 ～ 1.011) ), Low hemoglobin (OR = 0.924, CI = 0.899 ～ 0.950), and enlarged medulla (OR = 5.147, CI = 1.472 ～ 17.994) were independent risk factors for perioperative blood transfusion during admission. The incidence of pneumonia during the hospitalization period was higher in the transfusion group than in the non-transfusion group, and the difference was statistically significant (P <0.05). According to logistic regression analysis, the prediction model P = 1/1 + exp- (1.828 + 0.64 * age + 0.007 * intraoperative blood loss + 1.638 * medulling-0.079Hb), global Omnibus test for predicting model coefficients: χ2 = 75.985, P <0.001; Hosmer and Lemeshow test: χ2 = 10.975, P = 0.203. The accuracy of the regression model prediction was 79.3%, the critical value was 0.595, the sensitivity was 78.5%, and the specificity was 81.1%. The area under the ROC curve was 0.857 and 95% CI (0.801 ～ 0.912). Conclusions The elderly and the intraoperative bleeding are more , Hemodialysis, and hemoglobin at the time of admission are independent risk factors for perioperative blood transfusion in elderly femoral intertrochanteric fractures, and the prediction model established has certain risk assessment value and can be applied to the clinic.