Mature sheep receiving supplements of sodium chloride into the rumen were given intravenous infusions of arginine vasopressin at rates varying from 4-6-23 pmol/min (2-10 mU/min). Infusion of the hormone led to an increase in urine flow and to increases in the amounts of sodium and chloride excreted, the effect on flow was, however, the greater so that the osmolality of the urine fell during the infusions. In sheep given intravenous infusions of a hypertonic sodium chloride solution addition of vasopressin to the infusate led to the formation of a larger volume of urine containing a higher proportion of the infused salt load compared to when the salt solution alone was given. As before the effect on flow was the greater and hence the osmolality of the urine was lower when the hormone was given. In other experiments intravenous infusion of a hypertonic sodium chloride solution at rates providing 2-8 mmol NaCl/min led to increases in urine flow and increases in sodium and chloride excretion, the size of these increases being proportional to infusion rate. Plasma vasopressin levels markedly increased during these infusions, the levels seen being similar to those seen in sheep given vasopressin in amounts which increased both urine flow and electrolyte excretion. This suggests that during hypertonic salt loading vasopressin probably contributes directly to the increases in urine flow and the increases in electrolyte excretion which are seen. Further evidence in support of this was obtained in experiments in which a greater natriuretic response was seen in sheep given a hypertonic sodium chloride solution into the carotid artery as opposed to the given a hypertonic sodium chloride solution into the carotid artery as opposed to the jugular vein and where it was shown that plasma vasopressin levels were indeed higher when the solution was given into the artery.