Study design: Pilot study. Objectives: The aim of the study was to develop a neurophysiological method to diagnose the cranial as well as the caudal level of a complete thoracic spinal cord injury (SCI) with higher precision than today's protocols. Setting: SCI unit Karolinska University Hospital, Stockholm, Sweden. Methods: Bipolar needle electromyography was recorded in intercostal spaces of five patients with chronic, complete thoracic SCI. Tests were performed during rest, during voluntary activation and during activation of lower body spasticity. Magnetic resonance imaging (MRI) was performed in each patient according to a protocol optimized for imaging near metal implants. Results: Three distinct patterns were found in each patient. Above the lesion we found voluntary activated, normal motor unit potentials (MUPs). At the neurological level and a varying number of segments below, denervated intercostal segments with fibrillation potentials and positive sharp waves appeared. Below the neurological level, normal MUP activated in concert with lower body spasticity was found. The number of denervated segments showed a significant correlation to the length of spinal cord discontinuity on MRI (r¼0.97, Po0.05). Conclusion: Intercostal neurophysiology in combination with MRI optimized for imaging near metal implants can be used to determine the extent of a chronic complete thoracic SCI, both anatomically and functionally. The described method increases the sensitivity to detect delicate neurological changes related to the dynamic of the pathology that follows SCI and may be useful in analyzing outcome in clinical trials.