Objective To compare the differences of chronic obstructive pulmonary disease (COPD) models induced by smoke inhalation through nose-mouth plus LPS or smoke exposure through wholebody plus LPS in rats, providing a new model for COPD model construction. Methods 90 male SD rats were randomly divided into normal control group, wholebody exposure group and nose-mouth inhalation group, with 30 rats/group. The wholebody exposure group were exposed in a homemade smoke box where smoke contacted with whole body of rats , and the smoke nose-mouth inhalation group were inhaled with somke via nose-mouth only in a "quantitative smoking device". Animals in both groups were exposed to smoke once a day for 60mins/time for 8 weeks, and LPS (1mg/kg) was injected through the trachea on day 1, 7, 15 and 21, respectively, to induce the COPD model. The quality control of smoke generated by quantitative smoking devices included the verification of the stability and uniformity of the concentration of smoke particles and the size distribution of smoke particles. At 4, 6 and 8 weeks of modeling period, pulmonary function examination, the content of inflammatory factor interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α) in alveolar lavage fluid and histopathological examination were performed to compare the differences between the two modeling methods. Results Quantitative smoking devices could produce smoke with stable concentrations of 1.1mg/L (counted as total particles) and 0.1mg/L (counted as nicotine), respectively, with a median mass particle size of 0.693μm (in nicotine) and a GSD of 1.463. Compared with the whole body exposure group, the indexes of pulmonary function FEV0.2/FVC and pulmonary compliance (Cdyn) in the nose-mouth inhalation group decreased more significantly, and the airway resistance (Penh) increased more significantly. The levels of IL-6 and TNF-α in the alveolar lavage fluid of rats in the nose-mouth inhalation group were significantly increased at 6 weeks after modeling, while those indexs in the wholebody exposure group were increased at 8 weeks after modeling. The lesion severity of bronchial inflammation after modeling was similar between the oral-nasal inhalation group and the wholebody exposure group, but the lesion severity of emphysema was more serious in the oral-nasal inhalation group (the time when appeared statistic difference: oral-nasal inhalation vs. wholebody exposure = 6 weeks of modeling vs. 8 weeks of modeling). The mean linear intercept (MLI) in the oral-nasal inhalation group increased significantly at 4-8 weeks of modeling, and the mean alveolar numbers (MAN) decreased significantly at 6-8 weeks of modeling. MLI increased significantly and MAN decreased significantly in the whole body exposure group after 8 weeks of modeling. In the oral-nasal inhalation group, significant abnormal changes were observed in pulmonary function indexes (FEV0.2/FVC, Cdyn, Penh), cytokine levels in Bronchoalveolar lavage fluid (IL-6, TNF-α), and alveolar histopathological changes (bronchial severity and emphysema pathological score, MLI, MAN) after modeling. However, the coefficient of variation (CV%) of each index was significantly lower than that of the whole body exposure group. Conclusion LPS (1mg/kg) endotracheal drip combined with smoke wholebody exposured or via oral-nasal inhalation both could establish a typical rat COPD model. Rats Inhalated smoke via oral-nasal could shorten the modeling period. The model could be completed after 6 weeks of continuous modeling, presenting typical symptoms of chronic obstructive pulmonary disease (pulmonary ventilation dysfunction, broncho-lung chronic inflammatory infiltration, accompanied by emphysema), and the difference between individual model animals were smaller (vs smoke exposure, CV% values were smaller).