Objective:To establish an animal model of pneumonia for research on clinical prevention and treatment of bacterial pneumonia by infecting immunocompromised rats with drug-resistant Acinetobacter baumannii ( Ab) strains. Methods:Drug-resistant Ab strains were selected. Forty-eight SD rats were randomly divided into four groups: normal group, cyclophosphamide control group (intraperitoneal injection of 45 mg/kg cyclophosphamide), bacterial infection group (intratracheal instillation of 1.5×10 8 CFU Ab suspension), and bacterial infection+ immunosuppression group (intraperitoneal injection of 45 mg/kg cyclophosphamide+ intratracheal instillation of 1.5×10 8 CFU Ab suspension). Flow cytometry analysis was used to detect the proportion of CD4 + , CD8 + and NK cells in rat peripheral blood before as well as 3 d and 7 d after infection. A lung function meter was used to detect peak inspiratory flow (PIF), peak expiratory flow (PEF), tidal volume (Vt ) and forced expiratory volume in the second second/forced vital capacity (FEV 200/FVC) at 3 d and 7 d after modeling. ELISA was used to detect the levels of IL-6, TNF-α and IL-10 in the alveolar lavage fluid. HE staining was used to observe the morphology of rat lung tissues in each group. Bacterial loads in rat lung tissues were counted by bacterial culturing. Results:A decrease in voluntary activity was observed in rats in the cyclophosphamide control group, bacterial infection group and bacterial infection+ immunosuppression group after modeling. Lung rales could be heard in the bacterial infection group and bacterial infection+ immunosuppression group. Compared with the normal group, the cyclophosphamide control group showed decreased proportion of CD4 + and CD11b + NK cells and increased CD8 + cells in peripheral blood; the bacterial infection group showed decreased PIF, PEF, Vt and FEV 200/FVC, increased IL-6 and TNF-α levels and decreased IL-10 level in the alveolar lavage fluid, and higher bacterial load in lung tissues with mild widening of alveolar walls and inflammatory cell infiltration ( P<0.05, P<0.01). Compared with the cyclophosphamide control group and the bacterial infection group, the bacterial infection+ immunosuppression group showed a lower proportion of CD4 + cells and a higher proportion of CD8 + cells in rat peripheral blood, decreased PIF, PEF, Vt and FEV 200/FVC, increased IL-6 and TNF-α levels and decreased IL-10 level in alveolar lavage fluid, higher bacterial load in lung tissues, and aggravated widening of alveolar walls and inflammatory cell infiltration ( P<0.05, P<0.01). The proportion of CD11b + NK cells in peripheral blood of rats in the bacterial infection+ immunosuppression group was significantly lower than that in the bacterial infection group ( P<0.05, P<0.01). Conclusions:A bacterial pneumonia model was successfully constructed by infecting rats with Ab alone or in combination with cyclophosphamide immunosuppression. In the model constructed with Ab and cyclophosphamide immunosuppression, the rats had more severe pneumonia, which might be related to the reduced cellular immune function and the aggravated bacterial infection in rat lung tissues by cyclophosphamide.