Polyglutamine expansion mutations in specific proteins underlie the pathogenesis of a group of progressive neurodegenerative disorders, including Huntington's disease, spinal and bulbar muscular atrophy, dentatorubral-pallidoluysian atrophy, and several spinocerebellar ataxias. The different mutant proteins share ubiquitous expression and abnormal proteostasis, with misfolding and aggregation, but nevertheless evoke distinct patterns of neurodegeneration. This highlights the relevance of the full protein context where the polyglutamine expansion occurs, and suggests different interactions with the cellular proteostasis machinery. Molecular chaperones are key elements of the proteostasis machinery and therapeutic targets for neurodegeneration. Here we provide a focused review on Hsp90, Hsp70, and their cochaperones, and how their genetic or pharmacological modulation affects the proteostasis and disease phenotypes in cellular and animal models of polyglutamine disorders. The emerging picture is that, in principle, Hsp70 modulation may be more amenable for long-term treatment by promoting a more selective clearance of mutant proteins than Hsp90 modulation, which may further decrease the necessary wild-type counterparts. It seems, nevertheless, unlikely that a single Hsp70 modulator will benefit all polyglutamine diseases. Indeed, available data, together with insights from effects on tau and alpha-synuclein in models of Alzheimer's and Parkinson's diseases, indicates that Hsp70 modulators may lead to different effects on the proteostasis of different mutant and wild-type client proteins. Future studies should include the further development of isoform selective inhibitors, namely to avoid off-target effects on Hsp in the mitochondria, and their characterization in distinct polyglutamine disease models to account for client protein-specific differences.