We show that an applied electric field in which the polarity is reversed every 15 minutes can improve the outcome from severe, acute spinal cord injury in dogs. This study utilized naturally injured, neurologically complete paraplegic dogs as a model for human spinal cord injury. The recovery of paraplegic dogs treated with oscillating electric field stimulation (OFS) (approximately 500 to 600 microV/mm; n = 20) was compared with that of sham-treated animals (n = 14). Active and sham stimulators were fabricated in West Lafayette, Indiana. They were coded, randomized, sterilized, and packaged in Warsaw, Indiana, and returned to Purdue University for blinded surgical implantation. The stimulators were of a previously unpublished design and meet the requirements for phase I human clinical testing. All dogs were treated within 18 days of the onset of paraplegia. During the experimental applications, all received the highest standard of conventional management, including surgical decompression, spinal stabilization (if required), and acute administration of methylprednisolone sodium succinate. A radiologic and neurologic examination was performed on every dog entering the study, the latter consisting of standard reflex testing, urologic tests, urodynamic testing, tests for deep and superficial pain appreciation, proprioceptive placing of the hind limbs, ambulation, and evoked potential testing. Dogs were evaluated before and after surgery and at 6 weeks and 6 months after surgery. A greater proportion of experimentally treated dogs than of sham-treated animals showed improvement in every category of functional evaluation at both the 6-week and 6-month recheck, with no reverse trend. Statistical significance was not reached in comparisons of some individual categories of functional evaluation between sham-treated and OFS-treated dogs (ambulation, proprioceptive placing); an early trend towards significance was shown in others (deep pain), and significance was reached in evaluations of superficial pain appreciation. An average of all individual scores for all categories of blinded behavioral evaluation (combined neurologic score) was used to compare group outcomes. At the 6-month recheck period, the combined neurologic score of OFS-treated dogs was significantly better than that of control dogs (p = 0.047; Mann-Whitney, two-tailed).