OBJECTIVE We have previously shown that incomplete global cerebral ischemia complicated by dense acidosis produces a profound secondary deterioration of energy metabolism and cerebral blood flow. Antioxidant treatment only partially averts this deterioration, suggesting that parallel or sequential mechanisms are involved in cerebral ischemic injury. We tested the hypothesis that a novel competitive N-methyl-D-aspartate (NMDA) receptor antagonist GPI 3000 (GPI) ameliorates metabolic injury and that the effectiveness of the iron-chelator and antioxidant deferoxamine (DFO) is augmented by combined therapy with GPI after incomplete global cerebral ischemia. METHODS Anesthetized dogs were treated with 30 minutes of global incomplete cerebral ischemia. Preischemic plasma glucose was raised to approximately 500 mg/dL to exaggerate lactic acidosis. Brain ATP, phosphocreatine, and pH were measured by 31P MR spectroscopy for 180 minutes of reperfusion. Neurophysiological outcomes were assessed by evoked potential monitoring. Five groups were treated with either saline; 75 mg/kg DFO preischemia plus 75 mg/kg at reperfusion onset, followed by 27.5 mg/kg per hour for the remainder of reperfusion (DFO group); 25 mg/kg GPI pretreatment, followed by 5 mg/kg per hour (GPI-pre group); 25 mg/kg GPI at reperfusion, followed by 5 mg/kg per hour (GPI-post group); or DFO and GPI-pre at the same doses (Combined group). RESULTS Ischemic cerebral blood flow (microspheres: 5 to 8 mL/min per 100 g) was similar among the groups. End-ischemic pHi was also similar; 5.9 in saline, 6.1 in DFO, 6.2 in GPI-pre, 6.2 in Combined, and 6.1 in GPI-post groups. Progressive hypoperfusion was observed in all groups except Combined during reperfusion. Metabolic recovery was improved relative to saline in all drug-treated groups. Phosphocreatine recovery was improved in Combined compared with DFO and GPI-pre groups. Somatosensory evoked potential recovery was not observed in the saline group and incomplete in all treatment groups. At 60 and 90 minutes of reperfusion, DFO, GPI-pre, and Combined groups demonstrated improved recovery relative to the saline group. CONCLUSIONS Pretreatment and posttreatment with GPI ameliorated postischemic metabolic failure, suggesting that NMDA-mediated mechanisms are more important in global cerebral ischemia complicated by dense, acidosis than early studies indicated. Combined treatment with GPI and DFO improved cerebral blood flow during reperfusion and one indicator of energy recovery. These data support the hypothesis that parallel therapy aimed at antioxidant and antiexcitotoxic mechanisms of ischemic brain injury augment recovery compared with the individual agents.