Since uncontrolled production of beta-amyloid (Abeta) is considered a key seeding event underlying progression of Alzheimer's disease (AD), elimination of excessive Abeta and preventing its reaccumulation constitute the primary therapeutic goal in preventing and treating AD. To date, immunoneutralization has been the most effective strategy in removing pre-existing cerebral Abeta. Both active and systemic passive immunizations are known to reduce cerebral Abeta and improve memory in transgenic murine models of AD. However, active immunization is associated with adverse effects such as encephalitis with perivascular inflammation and hemorrhage, while passive immunization has the potential to disrupt cerebral vasculature that is laden with amyloid and exposed to high levels of antibody in the blood. Intriguingly, intracerebroventricular passive immunization established in the authors' laboratory circumvented these problems. The authors demonstrated that a single intracerebroventricular injection of anti-Abeta antibody reduced the cerebral Abeta burden and Abeta-related astrocytosis, retarded reaccumulation of Abeta and restored Abeta-induced depletion of presynaptic SNAP-25, for at least 1 month and reduced inflammatory reactions for 1 week in AD murine models without producing inflammation, microhemorrhage or systemic histotoxicity. These facts suggest that intracerebroventricular anti-Abeta may be a safe method for the rapid clearance of pre-existing Abeta and retarding reaccumulation of Abeta in AD. Intracerebroventricular administration via a catheter and reservoir, may be combined with the development of humanized monoclonal antibody against Abeta. Intraventricular shunts and ventriculostomy are frequently employed with acceptable risk-to-benefit ratios in the treatment of various brain disorders, while humanized antibodies are currently used in clinical trials of brain diseases such as multiple sclerosis and lymphoma.