Abstract: OBJECTIVE To investigate the drug resistance of elderly patients with pulmonary tuberculosis, analyze the risk factors for multi-drug resistance, construct and validate the nomogram prediction model for the elderly patients with pulmonary tuberculosis. METHODS A total of 502 elderly patients with pulmonary tuberculosis who were treated in Changsha Central Hospital from Jan. 2024 to May 2025 were enrolled in the study and were divided into the multi-drug resistant group with 64 cases and the fully sensitive group with 350 cases according to the result of antimicrobial susceptibility testing. The patients who were one-drug resistant, multidrug-resistant or pandrug-resistant were excluded. Multivariate logistic stepwise regression analysis was performed for the risk factors for the multi-drug resistance of the elderly patients with pulmonary tuberculosis. The nomogram model was established, and the predictive value of the model was evaluated by means of receiver operating characteristic (ROC) curves. RESULTS Among the elderly patients with pulmonary tuberculosis, the drug resistance rate was 30.28%(152/502), the one-drug resistance rate was 8.96%, the multidrug-resistant rate was 5.98%, the multi-drug resistance rate was 12.75%, and the pandrug-resistant rate was 2.59%. The patients of the multi-drug resistant group were elder than those of the fully sensitive group. The proportions of patients who were rural areas, workers, had diabetes mellitus history, history of contact with tuberculosis, retreatment, no less than 3 lung fields involved in both lungs, tuberculosis cavity, and tree bud signs were higher in the multi-drug resistant group than in the fully sensitive group (P＜0.05). Multivariate analysis indicated that the diabetes mellitus history(OR=2.109), residence place (rural area)(OR=1.883), history of contact with tuberculosis(OR=2.065), retreatment (OR=2.784) and tuberculosis cavity (OR=2.537) were the risk factors for the multi-drug resistance(P＜0.05). The Brier index of the nomogram model was 0.091. Hosmer-Lemeshow test and calibration curves demonstrated a good fit; the area under the curve (AUC) was 0.862(95%CI:0.811 to 0.913), with the sensitivity 93.18%, the specificity 67.54%. CONCLUSIONS The drug resistance rate of the elderly patients with pulmonary tuberculosis is high. The diabetes mellitus history, residence place (rural area), history of contact with tuberculosis, retreatment and tuberculosis cavity are the major risk factors for the multi-drug resistance. The nomogram model that is established based on the above factors exhibits favorable predictive efficiency.