1. We have tested the widely accepted hypothesis that resting-state bicarbonate secretion of gastric fundus mucosa is mediated by Cl(-)-HCO3- exchange in the apical membrane of surface epithelial cells (SECs). To this end, SECs of isolated fundus mucosa of Rana esculenta were punctured with double-barrelled microelectrodes to measure intracellular pH (pHi). 2. No significant pHi changes were observed in response to changing luminal HCO3- and/or Cl- concentrations. The change in pHi (delta pHi) in response to luminal chloride substitution averaged 0.00 +/- 0.01 pH units (mean +/- S.E.M.; n = 48), and did not change after blocking putative basolateral acid/base transporters which could have masked the pHi response. 3. On the other hand, pHi responded readily and reversibly to luminal perfusion with either low-pH (pH 2.5) solution (delta pHi = -0.36 +/- 0.05; n = 4; P < 0.01) or CO2-free HCO3- Ringer solution (delta pHi = +0.10 +/- 0.01; n = 29; P < 0.001). These observations demonstrate that the solution change was effective and complete within 1 min and show that the apical membrane of SECs is permeable to CO2. 4. The apical membrane of frog SECs could not be stained with an antibody against the C-terminal end of the mouse Cl(-)-HCO3- exchanger isoform AE2, although this antibody readily stained the basolateral membrane of the oxyntopeptic cells (OCs). 5. In conclusion, the presence of a Cl(-)-HCO3- exchanger in the apical membrane of SECs of frog gastric fundus mucosa in the resting state could not be confirmed, but other models of HCO3- secretion cannot be fully excluded. Observations from electrical measurements, favouring a model of conductive HCO3- secretion, point to the OCs rather than the SECs as a site of origin of HCO3- secretion.