Using flowing isotonic NaCl or flowing molar KCl electrodes, the magnitude of liquid junction potentials between the electrode tip and gastric juice was measured directly, and calculated from electrolyte concentrations in gastric juice according to standard formulas. Gastric transepithelial potential difference was obtained by subtracting the junction potential from the measured gastric potential difference. junction potentials by direct experiment correlated closely with calculated values. The magnitude of the junction potential was shown to depend on the level of gastric acidity as well as on the nature of the flowing intraluminal electrode. However, transepithelial potential difference was always the same with the two electrode solutions. It is impossible to measure junction potentials with an intraluminal KCl agar electrode, but they must exist because these electrodes did not accurately measure transepithelial potential difference when gastric acidity was high. Pentagastrin caused a transient decline in transepithelial potential difference; this was not observed in patients with pernicious anemia. We conclude that accurate measurement of transepithelial potential difference must include correction for liquid junction potentials, which are sizable when gastric acidity is high. This can best be done with a flowing intraluminal electrode.