In Pekin ducks adapted to salt water, 1Asp - 5Val -angiotensin II, 1Asp - 5Ile -angiotensin II and 1Asp - 5Ile -tetradecapeptide were applied intracerebroventricularly (I.C.V.) during steady-state conditions evoked by continuous intravenous loading with 200 mosmol kg-1 saline. Each of the angiotensin II (AII) analogues caused a dose-dependent antidiuresis with a concomitant rise in urine osmolality and electrolyte concentration. Antidiuresis was linearly correlated with plasma arginine vasotocin (AVT). The elevation of plasma AVT occurred rapidly during I.C.V. stimulation with AII and declined exponentially to the pre-stimulation level. Under conditions of salt loading with 1000 mosmol kg-1 saline in which the ducks excreted the salt and water by their supraorbital salt glands, AII applied I.C.V. in a concentration of 1 nmol ml-1, inhibited the NaCl excretion via the salt glands. Arterial blood pressure and heart rate increased after I.C.V. microperfusion with 1 nmol ml-1 AII. This was not due to leakage of I.C.V. AII into the circulation because systemic application of AII required a 100-fold higher dose to elicit similar effects. Respiration rate remained constant. Systemically applied AVT which produced plasma levels similar to, or greater than, those caused by centrally acting AII resulted in the same antidiuretic responses but did not mimic the circulatory effects of I.C.V. AII. Specific AVT antiserum, injected intravenously, totally suppressed the renal response to I.C.V. AII and reduced the rise in blood pressure and heart rate by more than 50%. The anterior part of the third ventricle was more sensitive than the posterior part in eliciting the antidiuretic responses to I.C.V. applied AII. The particular combination of effects on renal excretion, salt gland secretion and cardiovascular function of centrally applied AII in the duck supports the idea that AII plays a major role as a central modulator of volume homeostasis.