Since the identification in 1984 of the amyloid beta protein (Abeta) as the major component of senile plaques and cerebrovascular amyloid in Alzheimer's disease (AD) brains, it is well accepted that the production of this protein is a crucial factor in the pathogenesis of AD. Abeta is produced by cleavage from the amyloid precursor protein (APP) and can form fibrils in vivo and in vitro. The formation of these fibrils is influenced by proteins that are found in association with Abeta-containing lesions in the AD brain. Several of these proteins arise by an inflammatory response of the brain to Abeta production. The distribution of different isoforms of Abeta, varying at the C-terminus of the peptide, varies among the Abeta-containing lesions in AD brains. Such variations may have consequences for the pathogenesis of AD because the various Abeta isoforms differ in their capacity to form fibrils, and they have different toxic effects on neurons and vascular cells, respectively. The experimental data indicate that the pathogenesis of senile plaques is different from the generation of cerebrovascular amyloidosis. Summarizing models for either type of AD pathology are presented.