Acetylcholine-controlled exocrine secretion by pancreatic acini has been explained by two hypotheses. One suggests that NaCl secretion occurs by secondary active secretion as has been originally described for the rectal gland of Squalus acanthias. The other is based on a "push-pull" model whereby Cl- is extruded luminally and sequentially taken up basolaterally. In the former model Cl- uptake is coupled to Na+ and basolateral K+ conductances play a crucial role, in the latter model, Na+ uptake supposedly occurs via basolateral non-selective cation channels. The present whole-cell patch-clamp studies were designed to further explore the conductive properties of rat pancreatic acini. Pilot studies in approximately 300 cells revealed that viable cells usually had a membrane voltage (Vm) more hyperpolarized than -30 mV. In all further studies Vm had to meet this criterion. Under control conditions Vm was -49 +/- 1 mV (n = 149). The fractional K+ conductance (fK) was 0.13 +/- 0.1 (n = 49). Carbachol (CCH, 0.5 micromol/l) depolarized to -19 +/- 1.1 mV (n = 63) and increased the membrane conductance (Gm) by a factor of 2-3. In the seeming absence of Na+ [replacement by N-methyl-D-glucamine (NMDG+)] Vm hyperpolarized slowly to -59 +/- 2 mV (n = 90) and CCH still induced depolarizations to -24 +/- 2 mV (n = 34). The hyperpolarization induced by NMDG+ was accompanied by a fall in cytosolic pH by 0.4 units, and a very slow and slight increase in cytosolic Ca2+. fK increased to 0.34. The effect of NMDG+ on Vm was mimicked by the acidifying agents propionate and acetate (10 mmol/l) added to the bath. The present study suggests that fK makes a substantial contribution to Gm under control conditions. The NMDG+ experiments indicate that the non- selective cation conductance contributes little to Vm in the presence of CCH. Hence the present data in rat pancreatic acinar cells do not support the push-pull model.