The microRNA (miRNA) expression profile by qRT-PCR depends directly on the most appropriate normalization strategy adopted; however, currently there is no universally adequate reference gene. Therefore, this study aimed to determine, considering RNA-Seq results, the most adequate endogenous normalizer for use in the relative quantification of urine miRNAs from head and neck cancer patients, treated with cisplatin chemoradiotherapy. The massive sequencing was performed to identify the miRNAs differentially expressed between the group with cisplatin nephrotoxicity (n = 6) and the one without (n = 6). The candidate endogen normalizer was chosen according to four criteria: (1) the miRNA must be expressed in most samples; (2) the miRNA must have a fold change value between 0.99 and 1.01; (3) the miRNA must have a p-value ≥ 0.98; and (4) the miRNA must not be commented on by the final GeneGlobe (Qiagen, Hilden, Germany) analysis. Four miRNAs met all the criteria (hsa-miR-363-5p, hsa-miR-875-5p, hsa-miR-4302, and hsa-miR-6749-5p) and were selected for validation by qRT-PCR in a cohort of 49 patients (including the 12 sequencing participants). Only hsa-miR-875-5p was shown to be an adequate normalizer for the experimental condition under investigation, as it exhibited invariant expression between the two groups.