Objective Explore the risk factors of post-transplant diabetes mellitus (PTDM) after heart transplantation and construct a predictive model. Methods Retrospective analysis of clinical data of 152 patients who underwent heart transplantation surgery in our hospital from March 2022 to February 2025. Patients who develop PTDM 6 months after surgery are designated as the PTDM group, while those who do not develop PTDM are designated as the non PTDM group. Collect preoperative, intraoperative, and postoperative related indicators from two groups of patients. Screening variables through univariate analysis, determining independent risk factors through multivariate logistic regression analysis, analyzing the combined predictive value of various factors for PTDM after heart transplantation through receiver operating characteristic (ROC) curve analysis, and constructing an integral evaluation scheme for PTDM prediction model after heart transplantation. Results In preoperative data, there were significant differences in age and BMI between patients in the PTDM group and those in the non PTDM group, P＜0.05。 In the intraoperative data, there were no significant differences in surgical time, extracorporeal circulation time, blood transfusion volume, anticoagulant use, and vasoactive drug use between PTDM group patients and non PTDM group patients, P＞0.05。 In postoperative data, there were significant differences in cumulative glucocorticoid dose, tacrolimus blood concentration, and infection status between patients in the PTDM group and those in the non PTDM group, P＜0.05。 Multivariate logistic regression analysis showed that age BMI、 The cumulative dose of glucocorticoids, blood concentration of tacrolimus, and infection status are influencing factors for PTDM after heart transplantation, P＜0.05。 The AUC value of the joint prediction is 0.877, the optimal cutoff value is 2.322, the sensitivity is 75.56%, the specificity is 86.92%, P<0.05. Conclusion There are many risk factors for abnormal glucose metabolism after heart transplantation. The predictive model can effectively predict the risk of PTDM occurrence and provide reference for personalized intervention.