Spinal cord injury after successful operation of the thoraco-abdominal aorta is an unpredictable complication which negatively affects the patient's quality of life. The main cause of spinal cord injury has been reported to be peroxidation of lipids. Edaravone, a free radical scavenger, has been used in the acute phase of cerebral infarction to ameliorate the brain damage. The aim of the present study was to evaluate the protective effect of edaravone on the neurological and histological outcome, and to examine the method of its administration so as to obtain the better effect, using animal models with ischemic spinal cord. Three groups of rabbits underwent surgical exposure of the abdominal aorta that was clamped for 20min to achieve spinal cord ischemia. Group A (n=6, control group) was given no medication. In group B (n=6), edaravone (3mg/ml saline/kg body weight) was administered intravenously 30min after reperfusion. In group C (n=6), the same dose of edaravone was administered at 30min, 24h and 48h after reperfusion. Neurological status was clinically assessed, using Tarlov's score, at 24h, 48h and 1week after reperfusion. Somatosensory evoked potential was measured preoperatively, at 20min after ischemia, at 30min after reperfusion, and at 24h, 48h and 1week after operation. Spinal cord sections were examined histologically to determine the degree of neuronal damage given by ischemic-reperfusion. Group A presented paraplegia with marked neuronal necrosis. Groups B and C maintained better neurogical function than Group A (p<0.001), and Group C was much better than Group B (p<0.05). In the model rabbits with 20min of ischemia-reperfusion, systemic repetitious administration of edaravone was found to have a more protective effect than a single administration on the spinal cord neurons and glia cells both neurologically and histologically.