Methylprednisolone sodium succinate (MPSS) has been proposed as a first-line treatment for acute spinal cord injury (SCI). Its clinical use remains, however, controversial because of the modest benefits and numerous side-effects. We investigated if MPSS could protect spinal neurons and glia using an in vitro model of the rat spinal cord that enables recording reflexes, fictive locomotion and morphological analysis of damage. With this model, a differential lesion affecting mainly either neurons or glia can be produced via kainate-evoked excitotoxicity or application of a pathological medium (lacking O2 and glucose), respectively. MPSS (6-10 μM) applied for 24 h after 1-h pathological medium protected astrocytes and oligodendrocytes especially in the ventrolateral white matter. This effect was accompanied by the return of slow, alternating oscillations (elicited by NMDA and 5-hydroxytryptamine (5-HT)) reminiscent of a sluggish fictive locomotor pattern. MPSS was, however, unable to reverse even a moderate neuronal loss and the concomitant suppression of fictive locomotion evoked by kainate (0.1 mM; 1 h). These results suggest that MPSS could, at least in part, contrast damage to spinal glia induced by a dysmetabolic state (associated to oxygen and glucose deprivation) and facilitate reactivation of spinal networks. Conversely, when even a minority of neurons was damaged by excitotoxicity, MPSS did not protect them nor did it restore network function in the current experimental model.