SUMMARY1. The method of sequential centrifugation has been used to obtain fluid samples from both the renal papilla and inner medulla of the rat.2. Experiments were carried out on Brattleboro rats with hereditary hypothalamic diabetes insipidus (DI; homozygous recessive), and on their (heterozygous) litter-mates with normal quantities of neurohypophysial vasopressin. Initial classification of the animals by measurement of urine volume and osmolality was confirmed by post-mortem bio-assay of the pituitary glands, in thirty-five out of forty-seven animals.3. In rats with DI, urine osmolalities comparable to those of heterozygous rats were obtained after four daily injections of 1 u. Pitressin Tannate in Oil (PTO). Under these conditions, when dehydration was superimposed for 72 hr, urine osmolality did not increase markedly. In heterozygous rats dehydrated for the same period of time, urine osmolality increased by some twofold.4. In rats with DI, the administration of PTO induced a rise of both sodium and urea concentrations in renal fluids and in urine. Dehydration during PTO administration caused a further rise of urea concentrations only. Dehydration per se significantly raised urinary and renal fluid urea concentrations, but sodium concentrations did not rise.5. In heterozygous rats dehydration per se increased both urea and sodium concentrations in renal fluids and urine.6. The evidence is discussed that the action of vasopressin involves factors apart from increasing the permeability of the distal nephron to water and urea. t M.R.C. Scholar.