Background: This study sought to compare the incidence of late toxicities among survivors of head and neck cancer treated by definitive chemoradiation using patient-reported versus physician-reported methodologies. Methods: Two validated quality of life instruments, the University of Washington Quality of Life questionnaire (UW-QOL) and the Functional Assessment of Cancer Therapy for Head and Neck questionnaire (FACT-H&N) were administered to patients returning for follow-up after chemoradiation for head and neck cancer. Only patients who had been clinically without evidence of disease for greater than 6 months were sampled. Scores were compared to physician-reported toxicities, which were recorded blindly and independently after each patient visit, using the National Cancer Institute’s Common Toxicity Criteria (version 4.0). Two by two contingency tables were constructed to assess differences between patient- and physician-reported responses using Fisher’s exact test. Results: Fifty patients (35 male; 15 females) completed both instruments. While 78% of patients reported an inability “to swallow certain solid foods” and 70% reported difficulty to “swallow naturally and easily” using the self-reported UW-QOL and FACT-H&N surveys, respectively, the incidence of grade 2+ esophageal dysfunction as reported by physicians was only 48% (p < 0.001). While 60% of patients reported “too little saliva” and 50% acknowledged having problems with "voice quality and strength" using the UW-QOL and FACT-H&N surveys, only 20% and 30%, respectively, were scored as having grade 2+ xerostomia and laryngeal toxicity. Significant discordance was also observed between patient- and physician-reported toxicities with respect to the domains of appearance (p = 0.02), pain (p = 0.01), activity/energy (p < 0.001), and mood (p = 0.001). Conclusions: Late toxicities are frequently under-reported by physicians after chemoradiation for head and neck cancer. The true incidence and severity of these treatment-related toxicities, with respect to both functional and psychosocial impairment, among survivors may be better evaluated by patient-reported methods.