The 'peripheral sink' hypothesis in Alzheimer disease (AD) posits that removal of amyloid-β (Aβ) from the periphery should decrease brain levels of Aβ. In a recent study, Andrew Schumacher and colleagues tested this idea by administering the Aβ-degrading protease neprilysin intravenously to a transgenic mouse model of AD. Although the treatment led to depletion of plasma Aβ, brain levels of this peptide were unchanged, indicating that peripheral Aβ clearance may not be a viable therapeutic option in AD.