Spinocerebellar ataxias (SCA) are a group of inherited neurodegenerative diseases. The most common SCAs are caused by an expansion of the trinucleotide CAG, which produces a polyglutamine tract in their encoded proteins, leading to their aggregation. Patients suffering from these disorders display movement impairments, and no effective treatment exists, leading to death 10-15 years after disease onset. Until now, nearly all attempts to use traditional N-methyl-d-aspartate receptor (NMDAR) antagonists to treat neurodegenerative diseases have failed. This is because NMDARs are not only promoters of neuronal death, but are also essential for high brain function. Recently, it was discovered that the interaction between NMDAR and an ion channel (TRPM4) underlies several types of neurodegenerative disorders. We hypothesized that this molecular underpinning might occur in SCA. To evaluate the therapeutic potential of blockade of NMDAR/TRPM4 coupling in mitigating neuropathology, we expressed recombinant interface inhibitors that compete with endogenous NMDAR/TRPM4 binding in SCA2 and SCA3 mouse models. As a complementary therapeutic strategy, we evaluated the neuroprotective effects of small molecules identified in a structure-based computational compound screening from over 1.13 million candidates that block NMDAR/TRPM4 interaction. We analysed by immunohistochemistry the levels of neuroinflammation markers, aggregate deposition, and neuronal integrity. Motor behavior and neuropathology were also performed on transgenic SCA3 mice. Both approaches resulted in a pronounced reduction of neuronal damage, protein aggregation and neuroinflammation in the mouse models of SCA2 and SCA3. Overall, these results indicate that NMDAR/TRPM4 interaction interface inhibitors are an effective therapeutic strategy to treat polyglutamine disorders.