When the non-coding repeat expansion in the C9ORF72 gene was discovered to be the most frequent cause of frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) in 2011, this gene and its derived protein, C9ORF72, were completely unknown. The mutation appeared to produce both haploinsufficiency and gain-of-function effects in the form of aggregating expanded RNAs and dipeptide repeat proteins (DPRs). An unprecedented effort was then unleashed to decipher the pathogenic mechanisms and the functions of C9ORF72 in order to design therapies. A decade later, while the toxicity of accumulating gain-of-function products has been established and therapeutic strategies are being developed to target it, the contribution of the loss of function starts to appear more clearly. This article reviews the current knowledge about the C9ORF72 protein, how it is affected by the repeat expansion in models and patients, and what could be the contribution of its haploinsufficiency to the disease in light of the most recent findings. We suggest that these elements should be taken into consideration to refine future therapeutic strategies, compensating for the decrease of C9ORF72 or at least preventing a further reduction.