Metabolic tolerance of low intracellular pH (pH(i)) was studied in well-oxygenated, perfused, neonatal, rat cerebrocortical brain slices (350 microns thick) by inducing severe hypercapnia. In each of 17 separate experiments 80 brain slices (approximately 3.2 g wet weight) were suspended in an NMR tube, perfused with artificial CSF (ACSF), and studied at 4.7 T with 31P and 1H NMR spectroscopy. Spectra obtained every 5 min monitored relative concentrations of lactate or high-energy phosphate metabolites, from which pH(i) and extracellular pH were determined. Unperturbed slice preparations were metabolically stable for > 10 h, with no significant changes occurring in pHi, ATP, phosphocreatine (PCr), inorganic phosphate, or lactate. Different levels of hypercapnia were produced by sequentially perfusing slices with the following different ACSF batches, each having previously been equilibrated with a specific mixture of CO2 in oxygen: (a) 10% CO2, 15 min of perfusion; (b) 30% CO2, 15 min of perfusion; (c) 50% CO2, 15 min of perfusion; (d) 70% CO2, 30 min of perfusion; (e) 50% CO2, 15 min of perfusion; (f) 30% CO2, 15 min of perfusion; and (g) 10% CO2, 15 min of perfusion. At the completion of this protocol slices were again perfused with fresh ACSF that was equilibrated with a 95% O2/5% CO2 gas mixture. In each of five separate 1H and 31P experiments, brain slices were recovered within 2 h after termination of exposure to high CO2. The pHi was determined from measurements of the chemical shift difference between phosphoethanolamine and PCr, using a calibration curve obtained for our preparation.(ABSTRACT TRUNCATED AT 250 WORDS)