Cerebrovascular deposits of amyloid beta-peptides (A beta) leading to amyloid angiopathy is a prominent feature of Alzheimer's disease (AD). But cerebrovascular lesions in AD are usually neglected and the focus of most recent research efforts has been on genetic factors and the neurotoxicity of A beta. Recently we demonstrated that in vitro or in vivo application of A beta produced vascular endothelial damage and an inflammatory response through the production of free radicals. To probe the differential cerebrovascular actions of the major A beta-peptides used in AD research, we used an in vitro paradigm in which low concentrations of soluble A beta-peptides were introduced into a medium containing physiologically viable bovine cerebral arteries. The major endogenous C-terminal variants of A beta (A beta(1-40) or A beta(1-42)) induced characteristic features of endothelial dysfunction such as enhanced vasoconstriction and diminished vasodilation. The synthetic neurotoxic peptide A beta(25-35) had no significant effect on the vascular endothelium. This illustrates that the vascular actions of A beta are distinct from the previously reported neurotoxic properties of these peptides. The antioxidants superoxide dismutase and n-tertiary-butyl-alpha-phenylnitrone (PBN) protected the endothelium from A beta toxicity suggestive of an oxygen radical mediated phenomenon. If uncontrolled, the endothelial damage can trigger neurodegeneration in circumscribed regions by ischemic or inflammatory mechanisms. Drugs specifically targeting A beta(1-42) may not accord the desired protection against amyloid toxicity as A beta(1-40) of cerebral or peripheral origin can still induce deleterious effects of A beta. The possibility that A beta-induced vascular dysfunction may be an early event in the neurodegenerative process offers new avenues for therapy.