In the apical plasma membrane of rabbit gallbladder epithelium, hydrochlorothiazide (HCTZ), besides its inhibiting action on Na+-Cl- symport and some side-effects, opens a SITS-sensitive Cl- conductance (Gcl), resulting in depolarization of the membrane due to a cell-to-lumen Cl- backflux. Some previous indications suggest that GCl is someway related to the Cl-/HCO3- exchanger. Thus the actions on the apical membrane of HCTZ and two inhibitors of the exchanger mainly studied in erythrocytes, namely phlorizin and phenylglyoxal (PG), have been compared. The transapical Cl- influx was measured radiochemically and the anion exchange fraction identified. The apical and transepithelial membrane potentials (Vm, Vm), the apical/basolateral membrane resistance ratio (Rm/Rs) and the transepithelial resistance (Rep) were measured with conventional microelectrodes and the changes in apical electromotive force and conductance were calculated. It has been shown that: (1) 2.5 x 10(-4) mol/l HCTZ abolishes Cl-/HCO3- exchange in a few seconds, (2) exchanger inhibition by HCTZ is direct and not mediated by carbonic anhydrase inhibition, (3) 2 mmol/l phlorizin and 4 mmol/l PG also inhibit the exchanger in a few seconds or at least rapidly and in parallel activate a depolarization of 6-7 mV, like HCTZ, (4) dose/response curves of the three drugs for depolarization activation and anion exchange inhibition overlap, (5) depolarization time courses are similar for the three drugs, (6) a decrease in the Rm/Rs ratio occurs in the presence of the three drugs, with a significant change in apical electromotive force when the luminal Cl- concentration is reduced, all this indicating the appearance of a substantial, quantifiable GCl which is absent in the absence of incubation with the drugs, (7) GCl values are similar, regardless of the drug which generates them, (8) the effects of the three drugs are not additive, and (9) stilbenes and dipyridamole abolish GCl (but not DPC) as well as the basal intrinsic conductance of the exchanger. On this basis it is concluded that some inhibitors of the Cl-/HCO3- exchanger either turn it into an anion channel or, less probably, activate a parallel GCl, as a consequence of the exchanger inhibition.