Twenty hogs were administered the following procedures before, during, and after overdistraction of the spinal column at T5-T6: somatosensory (SEP) and neurogenic-motor evoked potentials (NMEPs), hydrogen clearance procedures, Stagnara wake-up tests, and aortic-injection of silastic plastic. To ensure that overdistraction was possible, a nonosseous, circumferential osteotomy was made at T5-T6 and distraction applied in one-ratchet increments using Harrington instrumentation. Overdistraction was maintained for 3, 5, 6, 10, 15, 20, 25, or 30 minutes. Results indicated that the duration of overdistraction, as represented by lost NMEPs, was always correlated with the animal's clinical status on wake-up test. If overdistraction was maintained more than 6 minutes, 100% of the animals demonstrated positive wake-up results; if maintained between 5 and 6 minutes, 75% demonstrated positive wake-up results; and if maintained less than 5 minutes, only 25% demonstrated positive wake-up results. Time-to-loss of the NMEPs and SEPs, after onset of overdistraction, fell within two groups: slow and fast. In the slow group, it required slightly more than 20 minutes (mean = 20.6) for the potentials to be lost, while in the fast-loss group data were lost in slightly less than 4 minutes (mean = 3.6). Blood flow studies and inspection of the spinal cord revealed that the mechanism of action for the slow group appeared to be ischemia of the spinal cord that extended several centimeters above and below the site of maximum distraction. In the fast-loss group, it appeared that gross structural damage, with some very localized ischemia, were the mechanisms of actions influencing the integrity of the spinal cord.(ABSTRACT TRUNCATED AT 250 WORDS)