It is well known that arginine vasopressin (AVP) produces up to a 40-fold increase (0.1 to 4.0 microL/min.cm2) in net water flux across the amphibian urinary bladder under an osmotic gradient (mucosal side 10% hypotonic). No AVP effect is observed when the gradient is in the opposite direction (serosal hypotonic). Similar asymmetrical behavior to osmotic gradients occurs in the frog corneal epithelium. This rectification phenomenon has not been satisfactorily explained. We measured net water fluxes in bladder sacs and confirmed that AVP has no effect when the serosal bath is hypotonic. We reasoned that the 'abnormal' serosal osmolarity was inducing changes in membrane water permeability, the very parameter being measured. Thus, we studied the effect of solution osmolarity on diffusional water flow (Jdw) across the frog bladder using 3H2O. As expected, AVP doubled Jdw (in either direction from 12 to 21 microL/min.cm2) when the serosal solution was iso-osmolar regardless of mucosal osmolarity. However, in the AVP-stimulated bladders, hypo-osmolarity of the serosal solution reduced Jdw by 42%, an effect that was reversible when normal osmolarity was re-established. Amphotericin B (instead of AVP) was used to irreversibly increase the permeability to water of the apical membrane. Under these conditions, basolateral hypotonicity also reversibly decreased Jdw by 32%, suggesting the basolateral membrane as the site where permeability is reduced. SEM and TEM of the tissue shows extreme swelling when it was exposed to serosal hypotonicity with or without AVP and typical surface morphology changes following hormone stimulation. We conclude that this swelling may initiate a signaling mechanism that reduces basolateral water permeability. These findings constitute evidence of basolateral water channel permeability regulation, which can also contribute to cell volume regulation.