Vasopressin increases the permeability of the total urinary bladder, an analogue of the mammalian renal collecting duct, to water and small solutes, especially the amide urea. We have observed that three general anesthetic agents of clinical importance, the gases methoxyflurane and halothane and the ultrashortacting barbiturate methohexital, reversibly inhibit vasopressin-stimulated water flow, but do not depress permeability to urea, or the the lipophilic solute diphenylhydantoin. In contrast to their effects in vasopressin-treated bladders, the anesthetics do not inhibit cyclic AMP-stimulated water flow, consistent with an effect on vasopressin-responsive adenylate cyclase. The selectivity of the anesthetic-induced depression of water flow suggests that separate adenylate cyclases and cyclic AMP pools may exist for control of water and urea permeabilities in to toad bladder. Furthermore, theophylline's usual stimulatory effect on water flow, but not its effect on urea permeability, was entirely abolished in methoxyflurane-treated bladders, suggesting that separate phosphodiesterases that control water and urea permeabilities are present as well. We conclude that the majority of water and urea transport takes place via separate pathways across the rate-limiting luminal membrane of the bladder cell, and that separate vasopressin-responsive cellular pools of cyclic AMP appear to control permeability to water and to urea.