Most cases of familial Alzheimer's disease are caused by mutations in PSEN1 (encoding presenilin-1). However, whether these missense mutations function through a loss- or gain-of-function mechanism is unclear. Previously, a gain-of-function mechanism was proposed on the basis of increased Aβ42/Aβ40 ratios in patients with PSEN1-mutant familial Alzheimer's disease. Now, Raymond Kelleher III, Jie Shen and colleagues report that Psen1 knock-in mice generated with either of two missense mutations from familial Alzheimer's disease show complete loss of presenilin-1 function due to loss of catalytic activity, confirming earlier in vitro results (Neuron 85, 967–981, 2015). The phenotypes of homozygous knock-in mice were indistinguishable from those of Psen1-null mice: perinatal lethality, developmental defects and impaired neurogenesis due to disruption of Notch signaling. Heterozygous mutants developed normally but had reduced Aβ production and increased Aβ42/Aβ40 ratios by 3 months of age and showed Alzheimer's disease–like age-dependent neurodegeneration and memory impairment. These results suggest that restoration of normal presenilin-1 function might be a promising therapeutic strategy for some patients with Alzheimer's disease.