Hereditary spastic paraplegia (HSP) confines a group of heterogeneous neurodegenerative disorders characterized by progressive spasticity and lower limb weakness. Age of onset is highly variable even in familial cases with known mutations suggesting that the disease is modulated by other yet unknown parameters. Although progressive gait disturbances, lower limb spasticity and extensor plantar responses are hallmarks of HSP these characteristics are also found in other neurodegenerative disorders, e.g. amytrophic lateral sclerosis (ALS). HSP has been linked to ALS and frontotemporal degeneration with motor neuron disease (FTD-MND), since TDP-43 positive inclusions have recently been found in an HSP subtype, and TDP-43 are found in abundance in pathological inclusions of both ALS and FTD-MND. Furthermore, ataxin-2 (encoded by the gene ATXN2), a polyglutamine containing protein elongated in spinocerebellar ataxia type 2, has been shown to be a modulator of TDP-43 induced toxicity in ALS animal and cell models. Finally, it has been shown that ATXN2 with non-pathogenic intermediate-length CAG/CAA repeat elongations (encoding the polyglutamine tract) is a genetic risk factor of ALS. Considering the similarities in the disease phenotype and the neuropathological link between ALS and HSP we hypothesized that intermediate-length CAG/CAA repeats in ATXN2 could be a modulator of HSP. We show that in a cohort of 181 HSP patients 4.9 % of the patients had intermediate-length CAG/CAA repeats in ATXN2 which was not significantly different from the frequencies in a Danish control cohort or in American and European control populations. However, the mean age of onset was significantly lower in HSP patients with intermediate-length CAG/CAA repeats in ATXN2 compared to patients with normal length repeats. Based on these results we conclude that ATXN2 is most likely not a risk factor of HSP, whereas it might serve as a modulator of age of onset.