1. Intracellular pH (pH1) was measured by microfluorimetry in secretory endpieces isolated from sheep parotid glands and loaded with the pH-sensitive fluoroprobe 2', 7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF). 2. Stimulation with 1 microM acetylcholine (ACh) caused a large, transient decrease in pH1 of 0.37 +/- 0.02 pH units followed by a slower recovery. The transient, which was reduced by 60% in the absence of HCO3-, could be attributed mainly to HCO3- efflux. During sustained stimulation, pH1 increased to a value that exceeded the resting value by 0.083 +/- 0.023 pH units after 20 min. 3. The anion channel blocker NPPB (0.1 mM) reduced the transient acidification in response to ACh by 48% and raised pH1 during sustained stimulation. Simultaneous application of NPPB and ACh accelerated the re-alkalinization following the initial acidification, indicating that NPPB inhibits HCO3- efflux. 4. The stilbene derivative H2DIDS (0.5 mM) reduced the transient acidification in response to ACh by 76% but caused a marked decrease in pH1 during sustained stimulation. Simultaneous application of H2DIDS and ACh slowed the re-alkalinization following the initial acidification, indicating that the main effect of H2DIDS was to inhibit HCO3- accumulation. 5. In the absence of HCO3-, the recovery from an acid load was unaffected by ACh stimulation. Acid extrusion, although dependent on Na+, was not inhibited by amiloride (1 mM), clonidine (1 mM) or H2DIDS (0.5 mM) and was therefore provisionally attributed to a Na(+)-H+ exchanger isoform other than NHE1 or NHE2. 6. In the presence of HCO3-, the rate of recovery from an acid load was reduced during ACh stimulation, probably as a result of the increased efflux of HCO3-. Acid extrusion was dependent on Na+ and was significantly inhibited by H2DIDS. 7. We conclude that ACh-evoked HCO3- secretion in the sheep parotid gland differs from that in many other salivary glands by being driven predominantly by basolateral Na(+)-HCO3- cotransport rather than by Na(+)-H+ exchange.