Previous studies have shown that magnesium salts and the noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonist, NPS 1506, attenuated short-term cognitive deficits and histopathological changes associated with traumatic brain injury (TBI). We evaluated the long-term effects of both therapies after brain trauma. Young adult rats were subjected to parasagittal fluid-percussion brain injury and received either MgSO(4) (125 micromol/400 g rat; n = 12) 15 min post-injury, NPS 1506 (1.15 mg/kg; n = 12) 15 min and 4 hr post-injury, or vehicle (n = 9) 15 min post-injury. Uninjured animals (sham) received vehicle (n = 10). Learning function in these animals was evaluated using a water maze paradigm 8 months after injury or sham treatment, and the brains were examined for cortical and hippocampal tissue loss. Compared to sham animals, injured vehicle-treated animals displayed a substantial learning dysfunction, indicated by an increased latency to find a hidden platform in the water maze (P < 0.001). No improvements in learning, however, were found for injured animals treated with NPS 1506 or MgSO(4). Injury induced >30% loss of tissue in the ipsilateral cortex in vehicle-treated animals that was not reduced in animals treated with either NPS 1506 or MgSO(4). Treatment with MgSO(4) significantly reduced progressive tissue loss in the hippocampus (P < 0.001). These findings are the first to demonstrate long-term neuroprotection of hippocampal tissue by an acute treatment in a TBI model. These data also show that the previously reported broad efficacy of MgSO(4) or NPS 1506 observed shortly after brain trauma could not be detected 8 months post-injury.