Objective To analyze influencing factors and construction of a nomogram predictive model for anastomotic leak after radical esophageal and gastroesophageal junction carcinoma. Methods The patients who underwent radical esophagectomy at Jinling Hospital affiliated to Nanjing University School of Medicine from January 2018 to June 2020 were selected. After screening for related variables using SPSS univariate and multivariate logistic regression analysis, the "nomogram" was used to predict the risk of anastomotic leak based on R language. The predicted effects were verified by the receiver operating characteristic (ROC) curves. Results A total of 468 patients with esophageal carcinoma were collected, including 354 (75.64%) males and 114 (24.35%) females with a mean age of 62.8±7.2 years. The tumor was mainly located in the middle or lower stage; 51 (10.90% ) patients had postoperative anastomotic leak. In univariate logistic regression analysis, age, BMI, tumor location, preoperative albumin, diabetes mellitus, anastomosis mode, anastomosis site, and CRP might be associated with anastomotic leak (P＜0.05). The above data suggested by multivariate logistic regression analysis illustrate that age, BMI, tumor location, diabetes mellitus, anastomosis mode, and CRP were independent risks of anastomotic leak (P＜0.05). The nomogram was constructed according to the results of multivariate logistic regression analysis. The area under the curve (AUC) of ROC curve was 0.803 showing that the actual observations agree well with the predicted results. In addition, the decision curve analysis concluded that the newly established nomogram was significant for clinical decision-making. Conclusion The predictive model of anastomotic leak after radical esophageal and gastroesophageal junction carcinoma has a good predictive effect and is critical for guiding clinical observation, early screening and prevention.