1. The diffusional permeability of collecting ducts to 22Na+ and 36Cl- was measured in rat papillae in vitro. 2. The permeability of the collecting duct to 36Cl- was 0.72 (s.e.m. = 0.01; n = 356) microns/sec which was significantly higher than the value of 0.51 (s.e.m. = 0.01; n = 356) microns/sec measured for 22Na+. 3. Collecting ducts in papillae taken from rats on a high sodium intake had a 22Na+ permeability of 0.63 (s.e.m. = 0.04; n = 53) microns/sec which was significantly higher than the value on a normal salt intake (0.50, s.e.m. = 0.04; n = 46 microns/sec). 4. When papillae from normal rats were studied in plasma taken from salt loaded rats, the 22Na+ permeability of 0.59 (s.e.m. = 0.04; n = 18) microns/sec was significantly higher than when incubated in plasma from normal rats (0.44, s.e.m. = 0.05; n = 12) microns/sec. 5. An extract of urine with natriuretic activity had no effect on 22Na+ permeability when tested in this system. 6. Adrenalectomy, PGE2, indomethacin and antidiuretic hormone had no significant effect on 22Na+ and 36Cl- permeability. 7. A substance exists in plasma from salt loaded animals that increases the permeability of collecting ducts to sodium. This effect could explain the component of the natriuresis that follows saline infusion which is independent of changes in glomerular filtration rate, aldosterone, or proximal tubule reabsorption.