The purpose of this study was to develop a model of unilateral cervical (C4-C5) spinal cord contusion injury in the rat and to characterize the functional and histological consequences following three injury levels using a new weight-drop spinal cord injury device. We evaluated forepaw/forelimb and hindlimb functions by: (1) a horizontal ladder beam measuring paw misplacements and slips; and (2) the forelimb preference test which measures the forelimb used for pushing off to rear, for support, and to land on after rearing. Rats with a mild spinal cord injury displayed primarily a forepaw deficit (forepaw misplacements) for 8 weeks after injury. Paw preference also improved after injury, but failed to reach control levels even after 12 weeks. These rats had damage primarily to the rubrospinal, spinocervicothalamic, and the uncrossed lateral corticospinal tracts in the dorsolateral funiculus a well as some loss of the lateral spinothalamic tracts in the lateral funiculus. Rats with a moderate injury had a prominent forepaw deficit still evident at 12 weeks after injury as well as a mild but not significant hindlimb deficit. Paw preference improved slightly 12 weeks. There was a larger lesion in the dorsolateral and lateral funiculi than in mildly injured rats which extended into the ventrolateral funiculi. There was a significant loss of gray matter compared to rats with a mild injury. Rats with a severe injury displayed significant forelimb and hindlimb deficits throughout the 12 week testing period compared to rats with a mild or moderate injury, and also had a more severe paw preference bias (90%). The lesion encompassed the entire dorsolateral, lateral and ventrolateral funiculi with some disruption of the ventral funiculus. There was more significant gray matter necrosis compared to rats with either a mild or moderate injury. Thus, the spinal cord injury device we used may be useful for studying graded cervical spinal cord injury in rats and potential treatments or interventions, because both the behavioral and histological effects are reproducible and consistent.