Human bipedal gait requires supraspinal control and gait is consequently severely impaired in most persons with spinal cord injury (SCI). Little is known of the contribution of lesion of specific descending pathways to the clinical manifestations of gait deficits. Here, we assessed transmission in descending pathways using imaging and electrophysiological techniques and correlated them with clinical measures of impaired gait in persons with SCI. Twenty-five persons with SCI participated in the study. Functional assessment of gait included the Walking Index for Spinal Cord Injury (WISCI), the Timed-Up and Go (TUG), the 6-Min Walking Test (6MWT), and the maximal treadmill gait speed. Balance was evaluated clinically by the Berg Balance Scale (BBS). The amplitude of tibialis anterior (TA) motor-evoked potentials (MEPs) at rest elicited by transcranial magnetic stimulation as a measure of corticospinal transmission showed a moderately good correlation with all clinical measures (r(2)~0.5), whereas the latency of the MEPs showed less good correlation (r(2)~0.35). Interestingly, the MEP amplitude was correlated to atrophy in the ventrolateral rather than the dorsolateral section of the spinal cord where the main part of the corticospinal tract is located. TA intramuscular coherence in the beta and gamma frequency range has been suggested to reflect corticospinal transmission and was, consistent with this, found to be correlated to atrophy in the dorsolateral and ventrolateral sections of the spinal cord. Coherence was found to correlate to all clinical measures to the same extent as the MEP amplitude. The latency and duration of medium-latency responses in the soleus muscle to galvanic stimulation as measures of vestibulospinal transmission showed very good correlation to BBS (r(2)=-0.8) and moderately good correlation to the assessments of gait function (r(2)~0.4). 6MWT and gait speed were correlated to atrophy of the lateral sections of the spinal cord bilaterally, whereas BBS was correlated to atrophy of both lateral and ventral sections of the spinal cord. No significant correlation was observed between the electrophysiological tests of corticospinal and vestibulospinal transmission. Combination of different electrophysiological and anatomical measures using best subset regression analysis revealed improved prediction of gait ability, especially in the case of WISCI. These findings illustrate that lesion of corticospinal and vestibulospinal pathways makes different contributions to impaired gait ability and balance following SCI and that no single electrophysiological or anatomical measure provide an optimal prediction of clinical gait and balance disability. We suggest using a combination of anatomical and electrophysiological measures when evaluating spinal cord integrity following SCI.