Abstract: Objective To observe the effects of adipose-derived mesenchymal stem cells (ADSCs) with continous over-expression of glial cell line-derived neurotrophic factor (GDNF) on the motor function recovery and nerve regeneration of sciatic nerve of rats after electrical injury. Methods Five SD rats were collected to prepare ADSCs with over-expression of GDNF. One hundred and fifty SD rats were divided into normal control group (N), GDNF-ADSCs group (GA), ADSCs group (A), GDNF group (G), and physiological saline group (P) according to the random number table, with 30 rats in each group. Rats in group N were routinely fed without treatment, and rats in the other 4 groups were inflicted with electrical injury on sciatic nerve of thigh of the right hind leg. Rats in groups GA, A, G, and P were respectively injected with 100 μL suspension of ADSCs with over-expression of GDNF (1×10⁷ cells per mL), 100 μL ADSCs suspension (1×10⁷ cells per mL), 100 μL GDNF solution (100 mg/L), and 100 μL physiological saline to the surface of the injured nerves immediately after injury. Six rats of each group were collected for measuring hind limb stride from post injury week (PIW) 1 to 8, and morphology of the sciatic nerves was observed in PIW 8. In PIW 4, the protein expression of GDNF of sciatic nerves of the rest rats in each group was determined with Western blotting. Data were processed with one-way analysis of variance, analysis of variance of repeated measurement, and SNK test. Results Compared with that of group N, the hind limb stride values in groups GA, A, G, and P were significantly lower at each time point (with P values below 0.05). Compared with those of group P, the hind limb stride values in group GA from PIW 3 to 8, in group A in PIW 3, 5, and 7, and in group G in PIW 3, 5, 7, and 8 were significantly longer (with P values below 0.05). The hind limb stride values in group GA from PIW 4 to 8 were respectively (10.83±0.97), (13.25±1.40), (12.86±1.42), (14.06±1.50), and (15.09±1.17) cm, which were significantly longer than those in group A [(8.90±0.82), (9.03±0.57), (9.27±0.36), (9.86±0.36), and (9.52±0.58) cm] and group G [(8.87±0.69), (8.51±1.18), (9.34±0.87), (9.76±0.67), and (9.50±1.22) cm], with P values below 0.05. Compared with that of group N, the number of myelinated nerve fibers of sciatic nerves was obviously decreased in group P but obviously increased in groups GA, A, and G; the diameter of axons was obviously shorter, and the myelin thickness was obviously increased in groups GA, A, G, and P in PIW 8 (with P values below 0.05). The number of myelinated nerve fibers in group GA was 31.2±0.8, which was significantly higher than that in group A (23.7±2.7), group G (22.3±2.7), or group P (9.3±2.8), with P values below 0.05. The diameter values of axons among groups P, A, G, and GA were similar (with P values above 0.05). The myelin thickness of rats in group GA was (3.41±0.34) μm, which was significantly thicker than that in group A [(2.64±0.37) μm] or group G [(2.41±0.34) μm], with P values below 0.05. In PIW 4, the protein expression of GDNF of sciatic nerves was significantly higher in groups P, A, G, and GA than in group N (with P values below 0.05), and the protein expression of GDNF in group GA was significantly higher than that in group P, A, or G (with P values below 0.05). Conclusions ADSCs over-expressing GDNF protein can obviously promote the motor function recovery and nerve regeneration of sciatic nerve of rats after electrical injury.