N-Ethylmaleimide (NEM) prevents the hydro-osmotic response to vasopressin. It has no significant effects on osmotic water movement in the absence of the hormone. The hydro-osmotic effects of cyclic AMP and hyperosmotic mannitol solutions are also reduced by NEM. After exposure to vasopressin, cyclic AMP, or hyperosmotic mannitol, the osmotic permeability declines to the preincubation level but after exposure to NEM this was prevented. This effect was seen after exposure to 10−4 and 10−5 m-NEM, i.e. concentrations that do not alter the onset of the response. The increased osmotic permeability after treatment with NEM is not due to a non-specific effect caused by a 'fixation' of the tissue in a distorted condition (such as occurs during rapid water transfer) since it was seen even after treatment of the tissue under conditions when no osmotic water movement was occurring, that is with iso-osmotic Ringer solution on both sides. The effect of NEM could not be mimicked by exposure of the bladder to 20 mm-Ca2+, 0·1 mm-Zn2+ or 1 mm-l-cysteine.