OBJECT In this prospective study, the authors offered protocol-selected patients a combination of parenteral steroids, decompression surgery, and localized cooling to preserve viable spinal cord tissue and enhance functional recovery. METHODS After acquiring informed consent, the authors offered this regimen with localized deep cord cooling (dural temperature 6°C) to 20 patients with a neurologically complete spinal cord injury to begin within 8 hours of injury. After decompression, the cord was locally cooled through the intact dura using a suspended extradural saddle at the site of injury for up to 4 hours, during which time spinal fusion was performed. Sensation and motor function were evaluated directly after the injury and again over a year later. The patients were evaluated using the 2011 amendment to the American Spinal Injury Association (ASIA) Impairment Scale. RESULTS Eighty percent of the 20 patients (12 with cervical and 4 thoracic injuries) with an initial neurologically complete cord injury had some recovery of sensory or motor function. All patients initially had ASIA Grade A impairment. Of 14 patients with quadriplegia, 5 remained ASIA Grade A, 5 improved to ASIA Grade B, 3 to ASIA Grade C, and 1 to ASIA Grade D. The remaining 6 patients had suffered a thoracic spinal cord injury, and of these 2 remained ASIA Grade A, 1 recovered to ASIA Grade B, 2 to ASIA Grade C, and 1 ASIA Grade D. All considered, of 20 patients, 35% remained ASIA Grade A, 30% improved to ASIA Grade B, and 25% to ASIA Grade C. Impairment in 2 (10%) of 20 patients improved to ASIA Grade D. The mean improvement in neurological level of injury in all patients was 1.05, the mean improvement in motor level was 1.7, and the mean improvement in sensory level was 2.8. Two patients recovered the ability to walk, 2 could extend their legs, 5 could sense bladder fullness, and 3 had partial ability to void voluntarily. Four males recovered subnormal ability to have voluntary erection sufficient for limited sexual activity. CONCLUSIONS The authors present here results of 20 patients with neurologically complete spinal cord injury treated with a combination of surgical decompression, glucocorticoid administration, and regional hypothermia. These patients experienced a better recovery than might have been expected had traditional forms of treatment been used. The benefit of steroid treatment for cord injury has been debated in the last decade, but the authors feel that research into the effects of cord cooling should be expanded. Given that the optimal neuroprotective temperature after acute trauma has not yet been defined, and may well be below that which is considered safely approachable through systemic cooling, methods that allow for the early attainment of such a temperature locally should be further explored. The results are encouraging enough to suggest the undertaking of controlled clinical trials of treatment using localized spinal cord cooling, where such treatment can be instituted within hours following injury.