An animal model of cauda equina syndrome was developed. Neurologic recovery was analyzed following immediate, early, and delayed decompression. Five experimental groups, each containing six dogs, were studied. Compression of the cauda equina was performed in all 30 dogs following an L6-7 laminectomy. The cauda equina was constricted by 75% in each group. The first group was constricted and immediately decompressed. The remaining groups were constricted for 1 hour, 6 hours, 24 hours, and 1 week, respectively, before being decompressed. Somatosensory evoked potentials were performed before and after surgery, before and immediately after decompression, and 6 weeks following decompression. Daily neurologic exams using the Tarlov grading scale were performed. At 6 weeks postdecompression, all dogs were killed, and the neural elements analyzed histologically. Following compression, all 30 dogs had significant lower extremity weakness, tail paralysis, and urinary incontinence. All dogs recovered significant motor function 6 weeks following decompression. The dogs with immediate decompression generally recovered neurologic function within 2-5 days. The dogs receiving 1-hour and 6-hour compression recovered within 5-7 days. The dogs receiving 24-hour compression remained paraparetic 5-7 days, with bladder dysfunction for 7-10 days and tail dysfunction persisting for 4 weeks. The dogs with compression for 1 week were paraparetic (Tarlov Grade 2 or 3) and incontinent during the duration of cauda equina compression. They recovered to walking by 1 week and Tarlov Grade 5 with bladder and tail control at the time of euthanasia. Immediately after compression, all five groups demonstrated at least 50% deterioration of the posterior tibial nerve evoked potential amplitudes. Six weeks after decompression, all five groups had a mean amplitude recovery of 20-30%. There were no statistical differences in recovery of somatosensory evoked potentials among the groups. Histologic analysis of the cauda equina in all groups demonstrated scattered wallerian degeneration and axonal regeneration. Areas of poor myelination, fibrosis, and macrophage activity were seen at the level of constriction. There were no significant differences in the histologic neuroanatomy of the five groups. It has been advocated that early decompression of cauda equina syndrome enhances neurologic recovery. This study does not support this premise. Although decompression allowed significant recovery in all 30 dogs, no significant differences were found in somatosensory evoked potentials, neurologic recovery, or histopathology in groups decompressed immediately, at 1 hour, 6 hours, 24 hours, or 1 week.