Phosphatic metabolite (perchloric acid extractable) concentrations of cerebral tissues were analyzed by phosphorus-31 nuclear magnetic resonance (P-31 NMR) spectfoscopy following external perfusion of the isolated rat brain (30 min or 60 min) under the following conditions: (a) constant perfusion pressure with either fluorocarbon-or erythrocyte-based medium, and (b) constant perfusate flow rate (3 ml/min) with the erythrocyte-based medium. Metabolite concentrations of control perfused brains were compared with those in nonperfused controls to provide a basis for detecting any qualitative or quantitative changes in cerebral metabolite composition. Metabolic responses of perfused brains to ischemia (incomplete ischemia, 83% reduction in flow for 10 min; transient complete ischemia for 1.5 or 2 min) were evaluated immediately after the ischemic episode and at selected time points during reperfusion (3 and 15 min). Alterations in cerebral metabolite levels induced by hypoxia were analyzed using a nonperfused rat brain model. Irrespective of the perfusion method employed, the phosphatic metabolites of control perfused rat brains were identical quantitatively to those of the nonperfused controls. Cerebral ischemia resulted in significantly increased levels of ADP, AMP + IMP, Pi, fructose 1,6-diphosphate, and glycerol 3-phosphate (global ischemia only), whereas ATP and phosphocreatine (PCr) levels declined significantly. The magnitude of these changes varied with the severity of the ischemia; however, following 15 min of control reperfusion metabolite levels had reverted to preischemic values. Significant perturbations in tissue phosphoethanolamine (3.848 resonance) content were evident at various time points during ischemia and postischemic recovery, which varied according to the perfusion conditions. In contrast to the changes observed in response to ischemia, hypoxia affected only cerebral high-energy phosphate levels. ATP and PCr levels were reduced, while a concomitant, essentially equimolar, increase in Pi and ADP was observed. The present studies indicate that in terms of phosphatic metabolites, the control equilibrated isolated perfused rat brain is quantitatively and qualitatively indistinguishable from the non-perfused rat brain in vivo regardless of the perfusion conditions (constant flow versus constant pressure). The metabolic responses to ischemia and hypoxia, as measured by P-31 NMR, were consistent with the pattern of changes reported elsewhere. Overall, P-31 NMR spectroscopic evaluation of the intact rat brain provides a potential experimental context for dynamic measures of cerebral metabolism under exogenously controlled conditions. The results reported herein from brain PCA extracts contribute a spectroscopic reference for understanding and interpreting the metabolic information contained in the P-31 profiles of the intact brain.