Background Concurrent chemoradiotherapy (CCRT) treatment incompletion is a known negative prognosticator for patients with head and neck cancer (HNC). Malnutrition is a common phenomenon which leads to treatment interruption in patients with HNC. We aimed to compare the performance of three nutritional tools in predicting treatment incompletion in patients with HNC undergoing definitive CCRT. Material and methods Three nutritional assessment tools, Mini Nutritional Assessment-Short Form (MNA-SF), Malnutritional Universal Screening Tool (MUST), and Nutritional Risk Screening 2002 (NRS-2002), were prospectively assessed prior to CCRT for HNC patients. Patients were stratified into either normal nutrition or malnourished groups using different nutrition tools. Treatment incompletion and treatment-related toxicities associated with CCRT were recorded. Results A total of 461 patients were included in the study; malnourished rates ranged from 31.0 to 51.0%. The CCRT incompletion rates were 4.9–6.3% and 14.5–18.2% for normal nutrition patients and malnourished patients, respectively. The tools had significant correlations with each other (Pearson correlation 0.801–0.837, p <0.001 for all) and accurately predicted the incompletion of CCRT. MNA-SF had the highest performance in predicting treatment-related toxicity, including emergency room visits, need for hospitalization, any grade III or higher hematological adverse events, and critical body weight loss, compared to the other tools. Conclusions MNA-SF, MUST, and NRS2002 were all shown to be competent tools for prediction of treatment incompletion and treatment-related toxicity in HNC patients undergoing CCRT. We suggest implementing nutritional assessment prior to treatment to improve the rate of treatment completion and to reduce treatment-related toxicity in HNC patients.