Head and neck squamous cell carcinoma (HNSCC), a major histologic subtype of head and neck cancer, presents great mortality and morbidity worldwide. The aim of this study is to discover new potential biomarkers closely correlated with HNSCC progression. In this study, weighted gene co-expression network analysis was applied to construct a co-expression network, and the brown module was identified as the most correlated with HNSCC progression. Hub gene identification combined with survival analyses determined RHCG as a candidate biomarker for cancer progression and prognosis prediction. Further experimental results proved that RHCG was aberrantly downregulated in HNSCC tissues and cell lines. Moreover, decreased RHCG expression was shown to be associated with advanced stage and dismal prognosis in HNSCC patients. Functional assays revealed that RHCG could inhibit cell viability, clonogenicity, cell migration in vitro and suppress tumor formation in vivo. Further bioinformatics study demonstrated that DNA promoter hypermethylation of RHCG could lead to its downregulation and serve as potential prognostic maker in HNSCC. Our study reveals that RHCG acts as a tumor suppressor gene that plays a crucial role in inhibiting tumorigenicity and metastasis in HNSCC, which will shed light on the potential diagnostic and therapeutic strategies for HNSCC.