AbstractHereditary ataxia (HA) is a heterogeneous group of complex neurological disorders, which represent a diagnostic challenge due to their diverse phenotypes and genetic etiologies. Next-generation sequencing (NGS) has revolutionized the field of neurogenetics, improving the identification of ataxia-associated genes. Notwithstanding, repeat expansions analysis remains a cornerstone in the diagnostic workflow of these diseases. Here we describe the molecular characterization of a consecutive single-center series of 70 patients with genetically uncharacterized HA. Patients’ samples were analyzed for known HA-associated repeat expansions as first tier and negative ones were analyzed by whole exome sequencing (WES) as second tier. Overall, we identified pathogenic/likely pathogenic variants in 40% (n = 28/70) and variants of unknown significance (VUS) in 20% (n = 14/70) of cases. In particular, 10 patients (14.3%, n = 10/70) presented pathogenic repeat expansions while 18 cases (30%, n = 18/60) harbored at least a single nucleotide variant (SNV) or a copy number variant (CNV) in HA or HSP-related genes. WES allowed assessing complex neurological diseases (i.e., leukodystrophies, cerebrotendinous xanthomatosis and atypical xeroderma pigmentosum), which are not usually referred as pure genetic ataxias. Our data suggests that the combined use of repeat expansion analysis and WES, coupled to detailed clinical phenotyping, is able to detect the molecular alteration underpinning ataxia in almost 50% cases, regardless of the hereditary pattern. Indeed, NGS-based tests are fundamental to acknowledge novel HA-associated genes useful to explain the remaining wide fraction of negative tests. Nowadays, this gap is problematic since these patients could not benefit from an etiological diagnosis of their disease that allows prognostic trajectories and prenatal/preimplantation diagnosis.