RNA-binding proteins (RBPs) interacting with target RNAs play essential roles in RNA metabolism at the post-transcription level. Perturbations of RBPs can accelerate cancer development and cause dysregulation of the immune cell function and activity leading to evade immune destruction of cancer cells. However, few studies have systematically analyzed the potential prognostic value and functions of RBPs in squamous cell carcinoma of head and neck (SCCHN). Here, for the first time, we comprehensively identified 92 differentially expressed RBPs from The Cancer Genome Atlas (TCGA) database. In the training set, a prognosis risk model was constructed with six RBPs, including NCBP2, MKRN3, MRPL47, AZGP1, IGF2BP2, and EZH2, and validated by the TCGA test set, the TCGA all set, and the GEO data set. In addition, the risk score was related to the clinical stage, T classification, and N classification. Furthermore, the high-risk score was significantly correlated with immunosuppression, and low expression of EZH2 and AZGP1 and high expression of IGF2BP2 were the main factors. Thus, the risk model may serve as a prognostic signature and offer highlights for individualized immunotherapy in SCCHN patients.