In order to provide a reproducible experimental spinal cord injury with immediate feedback on the mechanical properties of the impact, we have developed an electro-mechanically driven, feedback-controlled impaction device. The device is sensitive to the characteristics of the injured tissue and allows continuous manipulation of impact force or tissue displacement. The current study describes the anatomical and behavioral outcomes of a range of impacts and examines the ability of the device to produce a consistent traumatic injury. Rats were subjected to spinal cord trauma at the midthoracic level. Group II received a wide range of impacts that were preset at a desired force level and group III received impacts preset for a constant displacement of the spinal cord surface. Group I served as laminectomy controls. Behavioral testing included assessments of general and fine locomotor skills (open field and grid walking) and postural adjustment to displacement (inclined plane). Lesion volumes and percentage area of the cord occupied by the lesion were assessed postmortem. For group II, significant correlations between the physical descriptors of the impact and the behavioral measures were observed early during the postoperative period for open field and inclined plane performance and later in the recovery period for grid walking. Lesion measures correlated significantly with impact descriptors and with behavioral measures as well. Group III showed consistent behavioral deficits which partially recovered in the postoperative interval. The behavioral results correlated well with the lesion measures for this group also. The results indicate that it is possible to produce an intermediate lesion in the rat which results in consistent recovery with a residual deficit 3 weeks postoperatively, using a device that allows for immediate assessment of the physical descriptors of impact trauma.