Alzheimer’s disease is a major health problem with limited available medical treatment options. Immunotherapy is one approach with the potential to slow or reverse the disease process. In transgenic mouse models of amyloid deposition, anti-Aβ immunotherapy is remarkably effective at diminishing the amyloid burden and reversing the memory deficiency phenotype present in these mice. Three distinct mechanisms of antibody action have been proposed to mediate these benefits of anti-Aβ immunotherapy. The first is a catalytic dissolution of the Aβ fibrils, proposed by Beka Solomon and colleagues. A second mechanism is opsonization of the amyloid by the antibody and subsequent phagocytosis by macrophages proposed by Dale Schenk and the Elan group. A third mechanism proposed by DeMattos, Holtzman and colleagues is the peripheral sink hypothesis, arguing that circulating antibodies sequester Aβ and favor efflux of Aβ from the CNS over influx to the CNS. None of these mechanisms are mutually exclusive. Our research group has evaluated these mechanisms using intracranial injection and systemic administration of anti-Aβ antibodies. We found evidence supporting all three mechanisms, and suggest they may act synergistically to achieve the large effect size of the immunotherapeutic approach. However, in aged amyloid precursor protein transgenic mice administered anti-Aβ antibodies systemically, there is a redistribution of the amyloid from the parenchyma to vascular elements. This increase in congophilic angiopathy is associated with increased risk of microhemorrhage. Thus, although we favor continued testing of the immunotherapy to treat Alzheimer’s disease, we believe caution should be taken to minimize the risk of vascular leakage.