Testing of potential therapies for spinal cord injury has been significantly hampered by the unavailability of a standardized, reproducible animal model with predictable outcome at a given force of injury (dose-response). The rat was selected in the development of this model in preference to larger animals for economy and availability; this permits use of large numbers of animals to increase statistical validity. In the experiments reported in this article, a static load method (weight placed gently on cord) of inducing cord injury was evaluated. A total of 198 Sprague-Dawley rats were used. Under general anesthesia, a one-level laminectomy was carried out at T-12 with the dura mater intact. Weights varying from 80 to 150 g were lowered onto the dorsal surface of the intact dura mater for durations of 0 to 300 seconds. Recovery of motor function was assessed for up to 8 weeks using two behavioral tests, a modified Tarlov scale and an inclined plane test of hind limb motor function. A statistically significant relationship was found between force of injury and motor recovery as measured by the Tarlov scale, but this did not correlate with inclined plane performance; the duration that the weight rested on the cord did not influence outcome. Pathologically, there was variation in the extent of damage for a given injury load. A semiquantitative pathological assessment of cord injury showed a statistically significant correlation between pathological score and behavioral deficit as measured by the Tarlov scale, but this did not correlate with inclined plane performance. In view of these deficiencies, the static load technique does not seem to be an ideal model for spinal cord injury research.