Improved knowledge of the electrode-tissue impedance will be useful in optimizing the clinical protocols and resulting e�cacy of the existing and emerging approaches to spinal cord stimulation. Toward that end, the complex impedance (amplitude and phase) of in vivo ovine spinal cord tissue was measured at the electrode-pial subdural surface interface from 5 Hz to 1 MHz, and with the bi-polar electrodes oriented both parallel and perpendicular to the rostralcaudal axis of the spinal cord. At stimulation frequencies above 10 kHz, most of the impedance then becomes resistive in nature and the phase diference between the stimulation signal and the resulting current drops to ≈ 10˚, thus maximizing power transfer to the tissues. Also, at these higher frequencies, the current pulse maintains significantly greater fidelity to the shape of the stimulation signal applied across the electrodes. Lastly, there were lower impedances associated with parallel as opposed to perpendicular orientation of the electrodes, thus reflecting the efects of fiber orientation within the spinal cord. Impedance diferences of this kind have not been reported with epidural stimulation because of the electrical shunting efects of the intervening layer of relatively high conductivity cerebrospinal fluid. �ese observations provide a quantitative basis for improved models of spinal cord stimulation and suggest certain advantages for direct intradural stimulation relative to the standard epidural approaches.