BackgroundTo predict the 90-day mortality in patients with locally advanced head and neck squamous cell carcinoma (LAHNSCC) after concurrent chemoradiotherapy (CCRT), we propose a predictive scoring system, the Taipei Medical University (TMU)-CCRT Mortality Predictor Scoring System, and validate its accuracy. MethodsThis study extracted data of 16 029 patients with head and neck cancer who completed CCRT from the Taiwan Cancer Registry database (TCRD). Based on the 90-day mortality data after CCRT completion, the patients were divided into 2 groups: 90-day mortality (n = 1 068) and 90-day survival (n = 14 961). Stepwise multivariate Cox proportional hazards model was used to select the significant risk factors. ResultsMultivariate analysis revealed that age ≥ 50 or ≥ 70 years, pneumonia, sepsis, hemiplegia, moderate or severe renal disease, leukemia, and metastatic non-HNSCC solid cancers were significant risk factors affecting the 90-day mortality rate. Next, the risk scores of 0, 1 to 3, 4 to 6, and 7 to 9+ were categorized to indicated very low, low, moderate, and high risks (90-day mortality rate = 3.37%, 5.00%-10.98%, 16.15%-29.13%, and 33.93%-37.50%, respectively). ConclusionsOur TMU-CCRT Mortality Predictor Scoring System can accurately predict the 90-day mortality in CCRT-treated patients with LAHNSCC. Legal entity responsible for the study“Szu-Yuan Wu” without further recourse to the authors. FundingHas not received any funding. DisclosureThe author has declared no conflicts of interest.