A major problem in spinal cord injury research is quantification of motor function in animals. Most investigators in the field currently use neurologic scoring systems, relying on subjective observations of complex behaviours and assigning scores based on arbitrary criteria. These scoring scales are prone to observer bias and are nonspecific. We describe here a simple, reproducible, noninvasive, and objective test of a limited aspect of spinal motor function in cats, based on a well-known involuntary response of animals to sudden free fall. Free fall responses, or FFRs, have been studied in many species, including man, and are thought to be carried in ventral and lateral column pathways, i.e., vestibulospinal, reticulospinal, and rubrospinal tracts. We recorded the FFRs from hind and forelimb muscles of 100 cats before and after thoracic spinal cord injury. Hindlimb FFRs were shown to have three quantifiable components: a fast synchronous activation (E1) followed by a short silent period during which spinal segmental reflexes are inhibited (I1) and a late desynchronized excitatory burst (E2). Thoracic spinal injury produced hindlimb FFR losses ranging from greatly reduced amplitude to complete absence of response. Residual FFRs correlated with the extent of ventral column preservation and locomotory ability. Individual FFR components can be preserved. For example, some injured cats exhibited only 11 responses. Our work suggests that FFRs are a reliable and sensitive test of motor recovery in spinal cord injury.