By combining immunohistochemical technique with microassay methods, we analyzed regional energy metabolism in vulnerable and tolerant areas of gerbil brains during evolution of neuronal damage after bilateral common carotid artery occlusion for 10 min with subsequent reperfusion. Four animals were used for each reperfusion period. Based on the information from the immunohistochemical examination, we dissected out vulnerable and tolerant subregions of the hippocampus, cerebral cortex, and thalamus from freeze-dried 20-mthick sections, and measured the levels of creatine phosphate (P-Cr), adenine nucleotides, guanine nucleotides, and purine bodies by HPLC, and the levels of glucose, glycogen, and lactate by an enzyme-immobilized column method. There were no significant differences in the levels of metabolites between vulnerable and tolerant subregions of control brains. After reperfusion, both vulnerable and tolerant subregions recovered preischemic metabolic profiles by 2 days. Although the regional differences between vulnerable and tolerant subregions were minimal at each reperfusion period, there were delays in the recovery of P-Cr, ATP, and/or total adenine nucleotides in all vulnerable subregions. A decline of P-Cr, ATP, and GTP levels without change in %ATP, AMP, or purine bodies occurred after reperfusion for 3 days, coinciding with the development of immunohistochemical damage by the immunoreaction for microtubule-associated protein 1A. The results supported the notion that subtle but sustained impairment of energy metabolism caused by mitochondrial dysfunction in the early reperfusion period might trigger delayed neuronal death in vulnerable subregions. Key Words: Immunohistochemistry-Energy metabolism-Selective vulnerability-Microtubule-associated protein-Cerebral ischemia. J. Neurochem. 72, 1232Neurochem. 72, -1242Neurochem. 72, (1999.Since the report by Lowry et al. (1964) of rapid depletion of high-energy phosphates and glucose-related metabolites after decapitation, the depletion of highenergy phosphates, such as creatine phosphate (P-Cr) and ATP, and depletion of glucose and glycogen with accumulation of lactate have been considered to be the most sensitive and reliable metabolic indicator for ischemic stress (for review, see Siesjö and Wieloch, 1985), where prompt reestablishment of blood flow resulted in quick recovery of high-energy phosphates and glucoserelated metabolites, but a longer duration of ischemia affected such metabolic recovery (Ljunggren et al., 1974;Kobayashi et al., 1977). More recent reports have shown that a certain degree of decrease in high-energy phosphates, even without evidence of energy failure, is associated with postischemic damage (Pulsinelli and Duffy, 1983;Arai et al., 1986;Yasumoto et al., 1988). However, analyses of high-energy phosphates and glucoserelated metabolites without precise information on histopathological damages may be misleading, because of the tissue heterogeneity of the brain and the existence of selective tissue vulnerability to ischem...