Objective: To investigate whether desferrioxamine (DFO) can enhance the homing of bone marrow mesenchymal stem cells (BMSCs) and improve neovascularization in random flaps of rats. Methods: BMSCs and fibroblasts (FB) of luciferase transgenic Lewis rats were isolated and cultured. Forty 4-week-old Lewis male rats were used to form a 10 cm×3 cm rectangular flap on their back. The experimental animals were randomly divided into 4 groups with 10 rats in each group: in group A, 200 µL PBS were injected through retrobulbar venous plexus; in group B, 200 µL FB with a concentration of 1×10 6 cells/mL were injected; in group C, 200 µL BMSCs with a concentration of 1×10 6 cells/mL were injected; in group D, cells transplantation was the same as that in group C, after cells transplantation, DFO [100 mg/(kg·d)] were injected intraperitoneally for 7 days. On the 7th day after operation, the survival rate of flaps in each group was observed and calculated; the blood perfusion was observed by laser speckle imaging. Bioluminescence imaging was used to detect the distribution of transplanted cells in rats at 30 minutes and 1, 4, 7, and 14 days after operation. Immunofluorescence staining was performed at 7 days after operation to observe CD31 staining and count capillary density under 200-fold visual field and to detect the expressions of stromal cell derived factor 1 (SDF-1), epidermal growth factor (EGF), fibroblast growth factor (FGF), and Ki67. Transplanted BMSCs were labeled with luciferase antibody and observed by immunofluorescence staining whether they participated in the repair of injured tissues. Results: The necrosis boundary of ischemic flaps in each group was clear at 7 days after operation. The survival rate of flaps in groups C and D was significantly higher than that in groups A and B, and in group D than in group C ( P<0.05). Laser speckle imaging showed that the blood perfusion units of flaps in groups C and D was significantly higher than that in groups A and B, and in group D than in group C ( P<0.05). Bioluminescence imaging showed that BMSCs gradually migrated to the ischemia and hypoxia area and eventually distributed to the ischemic tissues. The photon signal of group D was significantly stronger than that of other groups at 14 days after operation ( P<0.05). CD31 immunofluorescence staining showed that capillary density in groups C and D was significantly higher than that in groups A and B, and in group D than in group C ( P<0.05). The expressions of SDF-1, EGF, FGF, and Ki67 in groups C and D were significantly stronger than those in groups A and B, and in group D than in group C. Luciferase-labeled BMSCs were expressed in the elastic layer of arteries, capillaries, and hair follicles at 7 days after transplantation. Conclusion: DFO can enhance the migration and homing of BMSCs to the hypoxic area of random flap, accelerate the differentiation of BMSCs in ischemic tissue, and improve the neovascularization of ischemic tissue.