In the brain, angiotensin II (Ang II) induces various effects such as blood pressure increase, the release of arginine vasopressin (AVP) and drinking behaviour. In the present study, we investigated the ability of the angiotensin II type-I (AT(1)) receptor antagonist, candesartan, administered peripherally, to block the central effects of Ang II. Experiments were performed in conscious rats instrumented with an intracerebroventricular (i.c.v.) cannula or a guide cannula into the paraventricular nucleus (PVN) and arterial and femoral catheters. Candesartan was administered intravenously (i.v.) at doses of 0.01, 0.1, 1 or 10 mg/kg. Controls received vehicle (0.05 N Na(2)CO(3)). The drinking response ( n=10-11 per group), the pressor response ( n=7-8) and the release of AVP into the circulation ( n=10-11) following i.c.v. Ang II (100 ng) were measured 0.5, 2, 4 and 24 h following i.v. drug application. Candesartan inhibited the central responses to i.c.v. injected Ang II dose- and time-dependently. At the highest dose (10 mg/kg), the drinking and pressor responses and the release of AVP in response to i.c.v. Ang II were completely blocked at 4 h and still markedly inhibited 24 h after the antagonist application (by 85%, 48% and 86%, respectively). The lowest dose of the antagonist was without effect. In a further experiment, the release of AVP induced by microinjection of Ang II (100 ng) into the PVN was determined before and 4 h after administration of vehicle or candesartan (1 mg/kg, i.v.). Candesartan completely blocked the AVP release into the circulation induced by Ang II microinjection into the PVN. Our results demonstrate that candesartan administered peripherally effectively inhibits responses mediated by AT(1) receptors localised in periventricular brain regions as well as inside the blood-brain-barrier.