There are indications that both familial amyotrophic lateral sclerosis (ALS) and sporadic ALS phenotype and prognosis are partly regulated by genetic and environmental factors, supporting the theory that ALS is a multifactorial disease. The aim of this article was to assess the role of ATXN2 intermediate length repeats in a large series of Italian and Sardinian ALS patients and controls carrying a pathogenetic C9ORF72 GGGGCC hexanucleotide repeat. A total of 1972 ALS cases were identified through the database of the Italian ALS Genetic consortium, a collaborative effort including 18 ALS centers throughout Italy. The study population included: (1) 276 Italian and 57 Sardinian ALS cases who carried the C9ORF72 expansion; (2) 1340 Italian and 299 Sardinian ALS cases not carrying the C9ORF72 expansion. A total of healthy 1043 controls were also assessed. Most Italian and Sardinian cases and controls were homozygous for 22/22 or 23/23 repeats or heterozygous for 22/23 repeats of the ATXN2 gene. ATXN2 intermediate length repeats alleles (≥28) were detected in 3 (0.6%) Italian ALS cases carrying the C9ORF72 expansion, in none of the Sardinian ALS cases carrying the expansion, in 60 (4.3%) Italian cases not carrying the expansion, and in 6 (2.0%) Sardinian ALS cases without C9ORF72 expansion. Intermediate length repeat alleles were found in 12 (1.5%) Italian controls and 1 (0.84%) Sardinian controls. Therefore, ALS patients with C9ORF72 expansion showed a lower frequency of ATXN2 polyQ intermediate length repeats than both controls (Italian cases, p = 0.137; Sardinian cases, p = 0.0001) and ALS patients without C9ORF72 expansion (Italian cases, p = 0.005; Sardinian cases, p = 0.178). In our large study on Italian and Sardinian ALS patients with C9ORF72 GGGGCC hexanucleotide repeat expansion, compared to age-, gender- and ethnic-matched controls, ATXN2 polyQ intermediate length does not represent a modifier of ALS risk, differently from non-C9ORF72 mutated patients.