The amyloid hypothesis has dominated the thinking in our attempts to understand, diagnose and develop drugs for Alzheimers disease (AD). This article presents a new hypothesis that takes into account the numerous familial AD (FAD) mutations in the amyloid precursor protein (APP) and its processing pathways, but suggests a new perspective beyond toxicity of forms of the amyloid β-peptide (Abgr;). Clearly, amyloid deposits are an invariable feature of AD. Moreover, although APP is normally processed to secreted and membrane-bound fragments, sAPPbgr; and CTFb, by BACE, and the latter is subsequently processed by γ-secretase to Abgr; and CTFγ, this pathway mostly yields Ab of 40 residues, and increases in the levels of the amyloidogenic 42-residue Abgr; (Abgr;42) are seen in the majority of the mutations linked to the disease. The resulting theory is that the disease is caused by amyloid toxicity, which impairs memory and triggers deposition of the microtubule associated protein, Tau, as neurofibrillary tangles. Nevertheless, a few exceptional FAD mutations and the presence of large amounts of amyloid deposits in a group of cognitively normal elderly patients suggest that the disease process is more complex. Indeed, it has been hard to demonstrate the toxicity of Abgr;42 and the actual target has been shifted to small oligomers of the peptide, named Ab derived diffusible ligands (ADDLs). Our hypothesis is that the disease is more complex and caused by a failure of APP metabolism or clearance, which simultaneously affects several other membrane proteins. Thus, a traffic jam is created by failure of important pathways such as γ-secretase processing of residual intramembrane domains released from the metabolism of multiple membrane proteins, which ultimately leads to a multiple system failure. In this theory, toxicity of Abgr;42 will only contribute partially, if at all, to neurodegeneration in AD. More significantly, this theory would predict that focussing on specific reagents such as γ-secretase inhibitors that hamper metabolism of APP, may initially show some beneficial effects on cognitive performance by elimination of acutely toxic ADDLs, but over the longer term may exacerbate the disease process by reducing membrane protein turnover.