Due to the GC-rich, repetitive nature of C9orf72 hexanucleotide repeat expansions, PCR based detection methods are challenging. Several limitations of PCR have been reported and overcoming these could help to define the pathogenic range. There is also a need to develop improved repeat-primed PCR assays which allow detection even in the presence of genomic variation around the repeat region. We have optimised PCR conditions for the C9orf72 hexanucleotide repeat expansion, using betaine as a co-solvent and specific cycling conditions, including slow ramping and a high denaturation temperature. We have developed a flanking assay, and repeat-primed PCR assays for both 3′ and 5′ ends of the repeat expansion, which when used together provide a robust strategy for detecting the presence or absence of expansions greater than ∼100 repeats, even in the presence of genomic variability at the 3′ end of the repeat. Using our assays, we have detected repeat expansions in 47/442 Scottish ALS patients. Furthermore, we recommend the combined use of these assays in a clinical diagnostic setting.