This study aimed to characterize the prognostic and therapeutic implications of disulfidptosis-related genes (DRGs) in head and neck squamous cell carcinoma (HNSCC). Using multi-omics data from the Cancer Genome Atlas (TCGA) (n = 500) and Gene Expression Omnibus (GEO) cohorts (n = 270), we performed unsupervised consensus clustering, functional enrichment, immune deconvolution, and survival analyses to identify DRG-driven molecular subtypes. A disulfidptosis score (DRGscore) was developed via principal component analysis of prognosis-associated genes and validated across immunotherapy cohorts. Key results revealed eight DRGs with prognostic significance (P < 0.05), including RAC1 and SLC7A11, which form interaction networks linked to redox regulation and immune evasion. Four DRGclusters stratified patients into subgroups with distinct survival outcomes (P = 0.002), immune profiles, and pathway activities. DRGscore effectively predicted survival (P < 0.001), immunotherapy response (anti-PD1/PD-L1 cohorts: P = 0.0099–0.018), and drug sensitivity (A443654 IC50 = 0.12 μM vs. AICAR = 8.3 μM). Mutational profiling identified TP53 and MUC16 as high-risk biomarkers. These findings establish DRGscore as a robust prognostic tool integrating disulfidptosis biology and immune contexture, enabling risk-stratified therapeutic strategies for HNSCC.