Abstract: OBJECTIVE To analyze the risk factors for pulmonary infection in the patients with esophageal cancer, establish the prediction model and analyze the predictive values. METHODS The clinical data were retrospectively collected from the 382 esophageal cancer patients who received radiochemotherapy in the First Affiliated Hospital of Nanyang Medical College from Mar 2018 to Apr 2022. The enrolled patients were divided into the infection group with 41 cases and the non-infection group with 341 cases according to the status of pulmonary infection. The distribution of pathogens isolated from the patients with infection was observed, univariate analysis and multivariate logistic regression analysis were performed for the risk factors for the pulmonary infection in the esophageal cancer patients, the prediction model was established based on the risk factors, the predictive values were analyzed by means of receiver operating characteristic (ROC) curves, and the changes of inflammatory factors were observed after the patients had the infection. RESULTS Totally 41 patients had pulmonary infection among the 382 esophageal cancer patients undergoing radiochemotherapy, with the incidence rate 10.73%. Totally 71 strains of pathogens were isolated, 48 (67.61%) of which were gram-negative bacteria. The result of multivariate logistic regression analysis showed that the no less than 60 years of age, long history of smoking and tumor location at upper thoracic segment/middle thoracic segment were independent risk factors for the pulmonary infection in the esophageal cancer patients undergoing radiochemotherapy. The above indexes were included in the Logistic regression analysis to establish the regression equation: Logit(P)=-4.509+age ≥60 years old ×1.367+long-term smoking history×1.628+tumor location at upper thoracic segment/middle thoracic segment×0.987. ROC curve analysis showed that when the Logit(P) was greater than 13.58, the area under curve (AUC) was 0.941, with the sensitivity 85.37%, the specificity 87.98%. The levels of interleukin-1β(IL-1β), tumor necrosis factor-α(TNF-α), C-reactive protein(CRP) and procalcitonin (PCT) of the infection group were higher than those of the non-infection group(P＜0.05). CONCLUSION The gram-negative bacteria are dominant among the pathogens isolated from the esophageal cancer patients with pulmonary infection. The risk factors include the age, long-term smoking history and tumor location. The prediction model that is established based on the risk factors has high predictive value. The pulmonary infection may lead to the rise of levels of inflammatory factors.