The Spinal Cord Stimulation (SCS) therapy has been used for several decades to relieve pain in people that suffer from chronic pain. However, only half of the patients treated with SCS experience significant pain relief only half of the time. Mathematical modeling of the nervous system and its responses to electrical stimulation has guided the design of more effective and more efficient therapies. In this study, we developed a computational model of SCS, including a cable model of a fiber that was adjusted using published data of the morphology and ion channel dynamics of spinal fibers. We quantified model responses at several positions within the spinal cord using a wide range of stimuli, and validated the model using electrophysiological recordings in anesthetized animals. Our model identified a molecular mechanism of ultra-slow inactivation of spinal fibers that has not been described before. These findings can have great impact in our understanding of spinal electrophysiology, and may be used to design more effective and/or efficient SCS modalities for the treatment of chronic pain.