Excitotoxicity triggered by over-stimulation of glutamatergic receptors is considered to be a major component of damage following acute spinal cord injury (SCI). Using an in vitro model of neonatal rat SCI caused by transient application (1h) of the glutamate agonist kainate (0.05–0.1mM) to produce limited excitotoxicity, the present study investigated whether riluzole, a drug inhibiting glutamate release and neuronal excitability, could prevent neuronal loss and protect locomotor patterns 24h later. Immunohistochemical analysis of neuronal and motoneuronal populations was associated with recording of fictive locomotion induced by neurochemicals or dorsal root stimuli. Riluzole (5μM; 24 h application) per se exerted strong and persistent neurodepressant effects on network synaptic transmission from which recovery was very slow. When continuously applied after kainate, riluzole partially reduced the number of pyknotic cells in the gray matter, although motoneurons remained vulnerable and no fictive locomotion was present. In further experiments, riluzole per se was applied for 3h (expected to coincide with kainate peak excitotoxicity) and washed out for 24h with full return of fictive locomotion. When this protocol was implemented after kainate, no efficient histological or functional recovery was observed. No additional benefit was detected even when riluzole was co-applied with kainate and continued for the following 3h. These results show that modest neuronal losses evoked by excitotoxicity have a severe impact on locomotor network function, and that they cannot be satisfactorily blocked by strong neurodepression with riluzole, suggesting the need for more effective pharmacological approaches.