When toads, Bufo marinus, were mildly dehydrated, the permeability to water of their urinary bladders increased, but blood samples placed directly on the wall of an isolated assay bladder did not contain enough arginine vasotocin (AVT) to induce the observed response. One explanation for this discrepancy is that AVT is secreted intermittently and rapidly cleared from the blood, even though the hydroosmotic response persists. The hydroosmotic response of bladders to pulsatile secretion of AVT was simulated by dipping isolated bladders for 1 min in Ringer's fluid with AVT followed by washout of hormone. Bladders developed the characteristic increase in membrane permeability to water several minutes after AVT had been removed from the serosal bathing solution. With short periods of bladder exposure to hormone, higher concentrations were required for triggering responses equivalent to those induced with continuous stimulation. This requirement was greater for vasopressin than it was for AVT. Studies with tritium-labeled vasopressin indicated that there is a "receptor reserve" for the hydroosmotic action of vasopressin in the toad bladder. The magnitude of this reserve is reduced as the duration of stimulation with hormone is diminished. It is suggested that pulsatile exposure of the bladder to AVT increases the specificity of the hormone for hydroosmotic receptors and minimizes the development of target organ resistance. The large receptor reserve in this tissue may be required to capture short bursts of AVT secreted into blood, although this remains to be demonstrated.