All individuals with Alzheimer's disease (AD) experience a progressive loss of cognitive function, resulting from a neurodegenerative process characterized by the deposition of ϐ-amyloid (Aϐ) in plaques and in the cerebrovasculature, and by the formation of neurofibrillary tangles in neurons. The cause of the neuronal death is unknown but it is thought to be linked in some way to the ϐ-amyloid precursor protein (APP), which is the source of the Aϐ that accumulates in the AD brain. There are two pieces of supporting data for this: first, APP is overexpressed in Down's syndrome, which leads to AD-like neuropathology by the age of 40 in virtually all affected individuals; secondly, specific point mutations in APP cause some forms of familial AD. Our laboratory has focused on a specific aspect of APP and its connection with the neuronal destruction seen in AD. We have hypothesized that AD results from a progressive dysfunction of APP. In addition, on the basis of recent data generated by our laboratory and others, we propose that in the normal brain a percentage of APP is present as an integral protein of the plasma membrane that mediates the transduction of extracellular signals into the cell via its Aϐ-containing C-terminal tail. In AD, accumulation of abnormal levels of the C-terminus in the neuron disturbs this signal-transduction function of APP, causing disorders in the cell-cycle machinery and consequent apoptosis. Here, we discuss the key findings that support this hypothesis, and discuss its therapeutic implications for AD.