Vascular dysfunction and inflammatory processes may be early events in the pathology of Alzheimer's disease (AD). Even though amyloid beta-peptides (Abeta) play a prominent role in the initiation and progression of cellular dysfunction in AD, the precise in vivo actions of various Abeta-peptides has not been established. The cerebrovascular actions of the major Abeta-peptides (1-40) and (1-42) in live animals were investigated using an open cranial window technique. We show here that the Abeta-peptides cause vascular lesions, especially in the arterioles. In one set of experiments, leukocytes and platelets were tagged with Rhodamine 6G, soluble Abeta(1-40) infused intravenously for 2 minutes, and the vasculature video recorded for 90 minutes. In a second set of experiments, soluble Abeta(1-40) infusion was followed 30 minutes later by an infusion of soluble Abeta(1-42) and the vasculature recorded for 90 minutes. Fluorescent and transmission electron microscopic examinations demonstrated the following cerebrovascular action of Abeta-peptides: endothelial cell damage, leukocyte adhesion, platelet activation, thrombus formation, impeded blood flow, and smooth muscle cell damage. The vascular disruption observed were similar to those observed in the brains of some AD patients and may represent the initial phase of a vascular inflammatory response associated with cerebral amyloid angiopathy. The combination of Abeta(1-40) and (1-42) produced significantly more vascular disruption than Abeta(1-40) alone. Oral administration of conjugated estrogens in ovariectomized female rats protected them from the deleterious actions of Abeta-peptides. The reported protective effect of estrogen against AD may be mediated in part through prevention of cerebrovascular Abeta toxicity.