Lysosomes are digestive organelles responsible for endocytosis and autophagy. Recently, the malignancy and invasiveness head and neck squamous cell carcinoma (HNSCC) has been increasingly studied with the role of lysosomes. A list of lysosome-related genes were obtained from MSigDB. A Spearman correlation and univariate Cox regression analyses combined with differential expression analysis were conducted to detect differentially expressed lysosome-related genes related to prognosis. The prediction of prognostic signature was evaluated by plotting survival curve, ROC, and by developing a nomogram. Immune subtypes, infiltration of immune cells, GSVA, TIDE, IC50 of common chemotherapy and targeted therapy, GO, and KEGG function enrichment analyses were carried out to explore the immune microenvironment of the signature. We constructed a lysosome-related prognostic signature that could function as an independent prognostic indicator for patients with HNSCC. High-risk patients were better suited to receive Doxorubicin, Mitomycin C, Pyrimethamine, anti-PD-L1 and anti-CTLA-4 immunotherapy, whereas low-risk patients had sensitivity to Lapatinib. GO functional enrichment analysis showed that prognostic features were strongly associated with epidermis-related functions (e.g., epidermal cell differentiation, epidermal development, and keratinization). In addition, a KEGG function enrichment analysis revealed a potential relationship between the risk assessment model and cardiomyopathy. We constructed a prognostic signature based on lysosome-related genes and successfully predicted the survival outcome of HNSCC patients, which not only provides potential guidance for personalized treatment but also provides a new idea for precision treatment of HNSCC.