OBJECTIVES: Clinical studies of metformin in certain cancer types have demonstrated decreases in incidence and improvements in therapeutic response. Preclinical evidence supports multiple potential molecular pathways through which metformin may impart beneficial effects which impede the development and growth of head and neck cancer (HNC). We investigated the impact of metformin exposure following a diagnosis of HNC on the risk of all-cause mortality in a population-based cohort of Italian patients. METHODS: We used the Italian Emilia-Romagna Regional (RER) longitudinal healthcare database to conduct a retrospective cohort study following approximately 4 million adults (>=18) from 2003 to 2011. The RER database captures de-identified, fully-linkable demographic, hospital discharge including ICD-9 diagnostic and procedure codes, outpatient pharmacy, and specialty data for all residents of the region. Resection status and metastases were used to stage the HNC patients based on surgical procedures and secondary malignancies. Cox proportional hazard methods were used to model the covariate-adjusted time-dependent medication exposure survival association to minimize potential immortal-time-bias. RESULTS: During the study period, we identified 7,872 patients diagnosed with HNC after which 708 (8.99%) were exposed at some point to metformin and 3,626 (46.1%) died during follow-up (median = 2.98 years). Among them, after adjusting for potential confounders, we found no significant association between exposure to metformin and reduced risk of all-cause mortality (HR=1.03, 95% CI:  0.84 – 1.26). CONCLUSIONS: Our large population-based cohort study failed to find an advantage in survival outcomes for HNC cancer patients exposed to metformin, despite in-vitro mechanistic viability. Further research will be critical in clearly defining the role of metformin in HNC cancer survival using similar modeling approaches.