Alzheimer's disease (AD) is the most common form of dementia and is associated with the accumulation of amyloid-β (Aβ), a peptide whose aggregation has been associated with neurotoxicity. Drugs targeting Aβ have shown great promise in 2D in vitro models and mouse models, yet preclinical and clinical trials for AD have been highly disappointing. We propose that current in vitro culture systems for discovering and developing AD drugs have significant limitations; specifically, that Aβ aggregation is vastly different in these 2D cultures carried out on flat plastic or glass substrates vs. in a 3D environment, such as brain tissue, where Aβ confinement very likely alters aggregation kinetics and thermodynamics. In this work, we identified attenuation of Aβ cytotoxicity in 3D hydrogel culture compared to 2D cell culture. We investigated Aβ structure and aggregation in solution vs. hydrogel using Transmission Electron Microscopy (TEM), Fluorescence Correlation Spectroscopy (FCS), and Thioflavin T (ThT) assays. Our results reveal that the equilibrium is shifted to stable β-sheet aggregates in hydrogels and away from the relatively unstable/unstructured presumed toxic oligomeric Aβ species in solution. Volume exclusion imparted by hydrogel confinement stabilizes unfolded, presumably toxic species, promoting stable extended β-sheet fibrils. These results, taken together with the many recent reports that 3D hydrogel cell cultures enable cell morphologies and epigenetic changes that are more similar to cells in vivo compared to 2D cultures, strongly suggest that AD drugs should be tested in 3D culture systems as a step along the development pathway towards new, more effective therapeutics.