Conscious Merino ewes were given an intravenous hypertonic sodium chloride load of 4 mmol . min-I for 100 min. This resulted in increases in urine flow, sodium and potassium excretion and plasma sodium concentration and osmolality. Urinary vasopressin output and solute-free water reabsorption increased and plasma renin activity declined. Renal plasma flow and glomerular filtration rate (GFR) rose, as did the solute clearance. The change in urinary osmolality was related to the initial urine osmolality such that when the initial urine osmolality was high the urine became more dilute, and vice versa. Tubular sodium reabsorption increased but the fractional reabsorption rate fell.It is suggested that the increase in GFR was at least partly due to the increase in AVP and that the electrolyte loss can be accounted for by the increase in GFR without necessarily involving AVP or other hormonal effects at the tubular level.The principal role of arginine vasopressin (AVP) in the sheep, as in other animals, appears to be the conservation of water, and the amounts needed to effect this control are small and have no apparent effect on electrolyte excretion [Brook, Radford and Stacy, 1968;Yesberg, Henderson and Budtz-Olsen, 1970]. On the other hand, the diuretic and electrolyte-excreting effects of infused AVP in normal, non-hydrated sheep are well established [Kinne, Macfarlane and Budtz-Olsen, 1961;Cross, Thornton and Tweddell, 1963;Ostwald, 1963;Gans, 1964;Scott and Morton, 1976). We have shown that AVP can increase glomerular filtration rate (GFR) in such sheep [Yesberg, Henderson and Budtz-Olsen, 1973] and have suggested that this could at least partly account for the increased electrolyte excretion as well as the diuresis [Yesberg, 1974]. These effects on water and electrolyte excretion and GFR have also been observed with the pressor analogues of AVP, lysine-vasopressin and ornithine-vasopressin [Yesberg, 1974] but not with the non-pressor analogue, desamino-D-arginine-vasopressin [Yesberg, Henderson and Budtz-Olsen, 1978]. It might be argued that these observations are responses to pharmacological amounts of the hormone, but Scott and Morton [1976] showed that intravenous hypertonic saline loading caused the release of endogenous AVP which was associated with, and probably the cause of, an increased water and salt excretion, and that the plasma AVP levels achieved were similar to those produced by an exogenous AVP infusion rate of 11-5 pmol. min ' which also caused increased water and salt excretion. In experiments involving one or 331