Patient survival from head and neck squamous cell carcinoma (HNSCC), the seventh most common cause of cancer, has not markedly improved in recent years despite the approval of targeted therapies and immunotherapy agents. Precision medicine approaches that seek to individualise therapy through the use of predictive biomarkers and stratification strategies offer opportunities to improve therapeutic success in HNSCC. To enable precision medicine of HNSCC, an understanding of the microenvironment that influences tumour growth and response to therapy is required alongside research tools that recapitulate the features of human tumours. In this review, we highlight the importance of the tumour microenvironment in HNSCC, with a focus on tumour hypoxia, and discuss the fidelity of patient-derived xenograft and organoids for modelling human HNSCC and response to therapy. We describe the benefits of patient-derived models over alternative preclinical models and their limitations in clinical relevance and how these impact their utility in precision medicine in HNSCC for the discovery of new therapeutic agents, as well as predictive biomarkers to identify patients' most likely to respond to therapy.