‘Amyloid’ is a general term describing protein aggregates of cross-β-sheet conformation and fibrillar morphology. Amyloid deposits are connected to several neurodegenerative disorders including Alzheimer’s, Parkinson’s, and Huntington’s diseases. Although amyloid aggregates share common structural features, amyloid-forming proteins themselves are quite diverse in size, sequence, structure and stability. In Alzheimer’s disease, for example, the intrinsically disordered peptide β-amyloid (Aβ) is a small proteolytic product of a normal precursor protein that spontaneously self-associates into soluble oligomers and insoluble fibrillar aggregates. When amyloid deposits containing Aβ were first linked to Alzheimer’s disease more than 30 years ago, there was great hope that research into the molecular basis for amyloid aggregation would rapidly lead to discovery of therapies to cure the disease. Reality has proven otherwise: now there are just a few compounds on the market that may slightly slow the rate of cognitive decline. In this seminar I will review some of my lab’s efforts at developing a molecular-level understanding of amyloid protein aggregation and at creating compounds that might be neuroprotective. The major focus will be on Aβ, but with short forays into other amyloidogenic proteins. We will explore the hypothesis that, whereas homotypic (self) association leads to amyloid aggregates, heterotypic (non-self) association of amyloidogenic proteins may regulate and inhibit amyloid fibril formation. We will also briefly discuss some reasons why translating basic science into the clinic has been so challenging.