Alzheimer's disease has been the focus of several drug discovery approaches by the pharmaceutical industry. Four drug candidates coming out of such efforts have recently failed in late-stage clinical trials for lack of efficacy or safety concerns. These drugs were designed based on the presently dominant scientific hypothesis for Alzheimer's disease called the 'amyloid hypothesis'. This editorial will briefly review the failure of these drugs and the effect of this on the amyloid hypothesis. Rather than accept the status quo, this editorial suggests a revised version of this hypothesis to reconcile data from recent drug failures. We propose a two-phase disease process; a first phase that is independent of amyloid and a second robust phase dependent on the amyloid cascade. Further validation of this revised hypothesis could aid future drug discovery for this devastating disease.