Abstract: OBJECTIVE To analyze drug resistance characteristics, drug resistance genes and high risk factors in patients with pulmonary tuberculosis, and to construct a prediction model of risk factors in pulmonary tuberculosis patients. METHODS The clinical data of 452 patients with pulmonary tuberculosis admitted to the Fifth People's Hospital of Ganzhou City from Jan 2019 to Dec 2022 were retrospectively collected as research subjects. According to the response of patients to anti-tuberculosis drugs, they were divided into the drug-resistant group (n=108) and the drug-sensitive group (n=344). Drug resistance characteristics, baseline data and tuberculosis resistance gene mutations in pulmonary tuberculosis patients were analyze. Multivariate logistic regression analysis was used to analyze the high risk factors of drug resistance of pulmonary tuberculosis, and the high risk factor model was established. Receiver operating characteristic (ROC) curve was drawn to analyze the predictive value of the high-risk factor model for the occurrence of drug resistance in pulmonary tuberculosis. RESULTS A total of the 452 patients with pulmonary tuberculosis, 108 were drug-resistant, with the resistance rate of 23.89%, which included the rates of single drug resistance, non-single drug resistance, multi-drug resistance and extensive drug resistance were 7.96%, 5.09%, 7.52% and 3.32%, respectively. Multivariate logistic regression analysis showed that high risk factors for tuberculosis resistance were combined with diabetes mellitus, six lung fields of pulmonary involvement, gene mutation of rifampicin-resistant rpoB and isoniazid resistant-related katG/inhA (P＜0.05). The high-risk factor model was established as follows: Logit (P)=-10.384+ combined with diabetes mellitus×0.588+ 6 lung fields of pulmonary involvement ×0.573+ rpoB gene mutation ×0.533+ katG/inhA gene mutation ×0.651, and the area under ROC curve (AUC) value of this Logit (P) was 0.799, with sensitivity of 76.85% and specificity of 70.06%, respectively. CONCLUSION Drug resistance rate of pulmonary tuberculosis is high, and combined with diabetes mellitus, six lung fields of pulmonary involvement, and mutation of rpoB and katG/inhA gene are high risk factors for the occurrence of drug resistance. High risk factor model built on this has good diagnostic value for drug resistance of pulmonary tuberculosis. Clinically, this study may help doctors be able to design and implement more effective treatments for patients with drug-resistant pulmonary tuberculosis.