1. The effects of oxygenated and hypoxic bicarbonate/CO2, 10 and 25 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)-buffered artificial cerebrospinal fluid (ACSF) have been studied in a rat brain slice preparation. Double-barreled pH-selective microelectrodes were used to measure intracellular pH (pH(i)) and membrane potential in dorsal vagal motoneurons (DVMs) and to measure extracellular pH (pH(c)) in the dorsal vagal motonucleus. 2. In bicarbonate ACSF, pH(i) averaged 7.24 +/- 0.05 (mean +/- SE, n = 21) and ranged from 6.86 to 7.79 pH units. pH, averaged 7.13 +/- 0.08 (n = 10). 3. On changing from oxygenated bicarbonate ACSF to either 10 or 25 mM HEPES ACSF, pH(i) decreased by 0.13-0.15 units, and the membrane depolarized by 10-11 mV. pH(c) also decreased in 10 mM HEPES ACSF (pH(e) 6.89 +/- 0.02, n = 8) but not in 25 mM HEPES ACSF (pH, 7.15 +/- 0.06, n = 3). In most neurons changes in pH(i) preceded changes in membrane potential. 4. In bicarbonate ACSF and in 25 mM HEPES ACSF, there was a significant linear relationship between prehypoxic pH(i) and the direction and amplitude of the hypoxia-induced membrane potential change (either an hyperpolarization or a depolarization). 5. In 10 mM HEPES ACSF, hypoxia always induced a depolarization; there was no correlation between prehypoxic pH(i) and the membrane potential response. 6. In bicarbonate ACSF and in 10 and 25 mM HEPES ACSF, hypoxia resulted in intracellular and extracellular acidification. However, the extracellular acidification in hypoxic 10 mM HEPES buffer was most pronounced (pH 6.40 +/- 0.11, n = 8), reflecting a preexisting extracellular acidification in oxygenated 10 mM HEPES buffer. 7. Various hypotheses that could give rise to a relationship between changes in membrane potential and pH are discussed; arguments are presented in favor of the concept that modulation of ion channels by either pH(i) or pH(e), or both, is responsible for the observed correlations.