Abstract: Objective To explore the effects of allogeneic bone marrow mesenchymal stem cells (BMSCs) on polarization of peritoneal macrophages isolated from rats with sepsis induced by endotoxin/lipopolysaccharide (LPS). Methods (1) BMSCs were isolated, cultured and purified from 5 SD rats with whole bone marrow adherent method. The third passage of cells were collected for morphologic observation, detection of expressions of stem cell surface markers CD29, CD44, CD45, and CD90 with flow cytometer, and identification of osteogenic and adipogenic differentiation. (2) Another 45 SD rats were divided into sham injury group (SI, n=5), LPS control group (LC, n=20), and BMSCs-treated group (BT, n=20) according to the random number table. Rats in groups LC and BT were injected with LPS (5 mg/kg) via tail vein to induce sepsis; rats in group SI were injected with the same amount of normal saline to simulate the damage. At post injury hour (PIH) 1, rats in group BT were given 1 mL BMSCs (2×10⁶/mL) via tail vein injection; rats in another two groups were injected with equal volume of phosphate buffer saline. Five rats in group SI at PIH 24 and in groups LC and BT at PIH 6, 12, 24, and 48 were sacrificed to harvest lung tissue for pathological observation with HE staining. In addition, rats in group SI at PIH 24 and in groups LC and BT at PIH 24 and 48 were simultaneously performed with intraperitoneal injection of low-glucose DMEM. Then peritoneal fluid was harvested to culture peritoneal macrophages. Flow cytometer was used to assess the positive expression of cell makers of macrophages including CD68 (making gate), CD11c, and CD206 in group SI at PIH 24 and in groups LC and BT at PIH 24 and 48. Data were processed with one-way analysis of variance and LSD test. Results (1) The third passage of cells showed uniform fiber-like shape similar to fibroblasts. These cells showed positive expressions of CD29, CD44, CD90 and weak positive expression of CD45. They were able to differentiate into osteoblasts and adipocytes. These cells were identified as BMSCs. (2) At PIH 24, the structure of pulmonary alveoli of rats in group SI was clear and complete with no congestion or inflammatory cell infiltration. At PIH 6, the structure of pulmonary alveoli of rats in groups LC and BT was clear with a small amount of inflammatory cell infiltration, slight congestion and pulmonary interstitial thickening. At PIH 12, the inflammatory responses in lung tissue of rats in group LC were more severe than those in group BT with a large amount of inflammatory cell infiltration, serious congestion, and obvious pulmonary interstitial thickening. The pathological results of rats in group BT at PIH 12 was consistent with the results at PIH 6. At PIH 24, the pathological results of rats in groups LC and BT were similar to the results at PIH 12. At PIH 48, the structure of pulmonary alveoli tissue of rats in group LC was still severely disrupted, with a large number of inflammatory cell infiltration and congestion in lung tissue, but pulmonary interstitial thickening was slightly alleviated than before. The condition of rats in group BT nearly recovered to that in group SI. (3) At PIH 24, the positive expression rate of CD11c in peritoneal macrophages of rats in group LC [(83±10)%] was close to that in group BT [(87±7)%, P>0.05], and they were both significantly higher than the rate in group SI [(55±12)%, with P values below 0.01]. The positive expression rate of CD11c in peritoneal macrophages of rats in group LC [(59±11)%] at PIH 48 was close to that in group SI at PIH 24 (P>0.05), and they were both significantly higher than the rate in group BT [(20±11)%] at PIH 48 (with P values below 0.01). At PIH 24, the positive expression percentages of CD206 in peritoneal macrophages of rats were similar among the three groups (with P values above 0.05). The positive expression percentage of CD206 in peritoneal macrophages of rats in group SI at PIH 24 was close to that in group BT at PIH 48 (P>0.05), and they were both significantly lower than the percentage in group LC at PIH 48 (with P values below 0.01). Conclusions BMSCs can reduce the pathological inflammatory responses in the lung of rats with sepsis and inhibit peritoneal macrophages from polarizing into M1 phenotype, whereas they can not promote macrophages to polarize into M2 phenotype.