Radiotherapy remains the standard treat74ment for head and neck squamous cell carcinoma (HNSCC), however, resistance remains a real clinical problem despite the improvements and development of new treatment strategies. Patient outcome could potentially be improved if the response to irradiation could be predicted allowing alternative treatments to be implemented. The current project has investigated the ability to maintain HNSCC tissue under pseudo in vivo-like conditions using a bespoke microfluidic device. The interrogation of the tissue with irradiation has also been performed with a view to predicting outcome and ultimately personalising medicinal treatment. Following extensive optimisation of key steps in device establishment, e.g. flow rates, serum concentration, oxygen perfusion, using rat liver and human tumour tissue, a series of HNSCC biopsies were divided and placed into parallel microfluidic devices and pieces of each tumour were subjected to single-dose irradiation (5, 10, 15, and 20Gy). Lactate dehydrogenase (LDH) release was measured in effluent collected every 2hr. In five of the tumours (primary n=3; metastatic lymph node n=2), frozen tissue sections were stained for cytokeratin (CK), M30 (detects cleaved CK-18), γH2AX, Ki-67 and for TUNEL. A CK18-labelling index (CK18-LI) was developed by expressing the percentage of apoptotic area (M30) over the total tumour area (CK+ staining). The positive γH2AX, Ki-67 expression and TUNEL were evaluated as positive nuclei within selected tumour areas. Single-dose irradiation induced variable, yet higher, CK18-LI compared with nonirradiated controls. In contrast no statistically significant differences in LDH release were observed between irradiated samples and controls. The percentage of Ki-67 expression reduced dose-dependently but not significantly following on-chip irradiation. Variation in expression profiles of these markers was identified between patients when using the same irradiation regimen demonstrating the potential of current microfluidic irradiation experimentation in monitoring patient’s radiotherapeutic response. In conclusion, microfluidics offers a potential tool in personalised medicine, capable of predicting a tumours response to irradiation prior to clinical administration although further validation experiments are required. The approach has equal applicability to all solid tumours and multiple types of treatment, i.e. chemo-radiation.