Glutathione is the major low-molecular weight thiol of eukaryotic cells. It is central to one of the two major NADPH-dependent reducing systems and is likely to play a role in combating oxidative stress, a process suggested to play a key role in Alzheimer's disease (AD). However, the nature and relevance of redox changes in the onset and progression of AD are still uncertain. Here, we combine genetically encoded redox sensors with our Drosophila models of amyloid-beta (Aβ) aggregation. We find that changes in glutathione redox potential (EGSH) closely correlate with disease onset and progression. We observe this redox imbalance specifically in neurons, but not in glia cells. EGSH changes and Aβ42 deposition are also accompanied by increased JNK stress signaling. Furthermore, pharmacologic and genetic manipulation of glutathione synthesis modulates Aβ42-mediated neurotoxicity, suggesting a causal relationship between disturbed glutathione redox homeostasis and early AD pathology.