〔Abstract〕 Objective To construct and validate a clinical diagnostic model to differentiate between pulmonary tuberculosis and non-tuberculous lung diseases. Methods Information was collected from 258 patients with respiratory system diseases, and they were divided into a training set of 152 cases and a test set of 106 cases with a ratio of 6 ∶4 using the random number seed method in R software. The training set was further divided into a tuberculosis group of 95 cases and a non-tuberculosis group of 57 cases, and the test set into a tuberculosis group of 65 cases and a non-tuberculosis group of 41 cases based on the diagnosis of pulmonary tuberculosis. A diagnostic model was constructed using multivariate logistic regression analysis to determine the influencing factors of pulmonary tuberculosis. The diagnostic value and clinical utility of the model were assessed using the receiver operating characteristic(ROC) curve, calibration curve, and decision curve analysis(DCA). Results CD161+%(OR=0.768; 95%CI0.697-0.845;P<0.001), AST(OR=0.961; 95%CI0.930-0.993;P=0.019), and smoking history(OR=3.181; 95%CI1.149-8.804;P=0.026) were identified as independent risk factors for the occurrence of pulmonary tuberculosis. In both the training and test sets, the area under the ROC curve(AUC) reached 0.870(95%CI0.816-0.924) and 0.887(95%CI0.827-0.948), respectively. The Hosmer-Lemeshow goodness-of-fit test showed a good fit(training set χ2=6.213,P=0.623; test set χ2=6.197,P=0.625). DCA indicated that the model had good reference significance for the diagnosis of the probability of pulmonary tuberculosis occurrence. Conclusion The diagnostic model constructed using the percentage of CD161+%, AST levels, and smoking history has certain diagnostic performance, facilitating rapid clinical differentiation between pulmonary tuberculosis and non-tuberculous lung diseases.