The polyglutamine (polyQ) diseases, including Huntington's disease and spinocerebellar ataxias, are classified as the protein misfolding neurodegenerative diseases like Alzheimer's and Parkinson's diseases, and they are caused by an abnormal expansion of the polyQ stretch in disease-causative proteins. Expanded polyQ stretches have been shown to undergo a conformational transition to a beta-sheet-dominant structure, leading to assembly of the host proteins into insoluble beta-sheet-rich amyloid fibrillar aggregates and their subsequent accumulation as inclusion bodies in affected neurons, eventually resulting in neurodegeneration. Based on cytotoxicity of the soluble beta-sheet monomer of the expanded polyQ protein, we propose the "Exposed beta-sheet hypothesis", in which both the toxic beta-sheet conformational transition and misassembly into amyloid fibrils of the disease-causative proteins contribute to the pathogenesis of the polyQ diseases, and possibly the other protein misfolding neurodegenerative diseases. Among the various therapeutic targets, the toxic conformational changes and aggregation of the expanded polyQ proteins are most ideal since they are the earliest events in the pathogenic cascade, and therapeutic approaches using molecular chaperones, intrabodies, peptides, and small chemical compounds have been developed to date. Furthermore, high-throughput screening approaches to identify polyQ aggregate inhibitors are in progress. We hope that protein aggregate inhibitors which are widely effective not only for the polyQ diseases, but also for many neurodegenerative diseases will be discovered in the near future.