L ow-protein diets cause a urinary concentrating defect in rats and humans. Previously, we showed that feeding rats a low (8%) protein diet induces a change in urea transport in initial inner medullary collecting ducts (IMCDs) which could contribute to the concentrating defect. Now, we test whether decreased osmotic water permeability (Pf) contributes to the concentrating defect by measuring Pf in perfused initial and terminal IMCDs from rats fed 18 or 8% protein for 2 wk. In terminal IMCDs, arginine vasopressin (AVP)-stimulated osmotic water permeability was significantly reduced in rats fed 8% protein compared to rats fed 18% protein. In initial IMCDs, AVP-stimulated osmotic water permeability was unaffected by dietary protein. Thus, AVP-stimulated osmotic water permeability is significantly reduced in terminal IMCDs but not in initial IMCDs. Next, we determined if the amount of immunoreactive aquaporin-2 (AQP2, the AVP-regulated water channel) or AQP3 protein was altered. Protein was isolated from base or tip regions of rat inner medulla and Western analysis performed using polyclonal antibodies to rat AQP2 or AQP3 (courtesy of Dr. M.A. Knepper, National Institutes of Health, Bethesda, MD). In rats fed 8% protein (compared to rats fed 18% protein): (a) AQP2 decreases significantly in both membrane and vesicle fractions from the tip; (b) AQP2 is unchanged in the base; and (c) AQP3 is unchanged. Together, the results suggest that the decrease in AVP-stimulated osmotic water permeability results, at least in part, in the decrease in AQP2 protein. We conclude that water reabsorption, like urea reabsorption, responds to dietary protein restriction in a manner that would limit urine concentrating capacity.