The amyloid-beta peptide (Abeta), implicated in the pathogenesis of Alzheimer's disease (AD), is produced through sequential proteolysis of the Abeta precursor protein (APP) by beta- and gamma-secretases. Thus, blocking either of these two proteases, directly or indirectly, is potentially worthwhile toward developing AD therapeutics. beta-Secretase is a membrane-tethered pepsin-like aspartyl protease suitable for structure-based design, whereas gamma-secretase is an unusual, heterotetrameric membrane-embedded aspartyl protease. While gamma-secretase inhibitors entered clinical trials first due to their superior pharmacological properties (for example, brain penetration) over beta-secretase inhibitors, it has since become clear that gamma-secretase inhibitors can cause mechanism-based toxicities owing to interference with the proteolysis of another gamma-secretase substrate, the Notch receptor. Strategies for targeting Abeta production at the gamma-secretase level without blocking Notch signalling will be discussed. Other strategies utilizing cell-based screening have led to the identification of novel Abeta lowering agents that likewise leave Notch proteolysis intact. The mechanism by which these agents lower Abeta is unknown, but these compounds may ultimately reveal new targets for AD therapeutics.