Purpose The aim of the study was to test the hypothesis that head and neck cancer (HNC) patients benefit from specialized follow-up (FU), as this strategy ensures timely detection of relapses for successful salvage treatment. This was done by evaluation of the pattern of failure, the temporal distribution of recurrences, and the outcome of salvage treatment in a contemporary cohort of HNC patients. Methods The study evaluated a cohort of 2062 consecutive patients treated with curative intent at Aarhus University Hospital from 1 January 2000 to 31 December 2013. The database of DAHANCA contained recordings of recurrent disease in 567 patients with primary tumors of the larynx, pharynx, oral cavity, nasal cavity, paranasal sinuses and salivary glands. A review of medical records was performed in order to update and supplement the database. Results Failures of the 567 patients were primarily in T-site (65%) followed by N-site (36%) and M-site (22%). The vast majority of the first recurrences occurred within the first years after primary treatment; 62%, 82%, and 91% within the first, second and third year, respectively. Totally, 51% were amenable for salvage treatment, and 44% benefited from salvage in terms of a complete response. Permanent tumor control was observed in 128 patients (23%) after one or two salvage attempts. The highest salvage rate was recorded in patients with primary glottic carcinoma (41%) and the lowest among hypopharyngeal cancers (2%). Asymptomatic recurrence was recorded in 12% of all recurrences and this was found to be a positive prognostic factor for disease-specific survival, as they had significantly better outcome after salvage. Conclusion Our data support the usefulness of specialized FU in terms of early detection of recurrent disease. In particular patients with silent recurrences benefited from early detection, as they had a significantly lower risk ratio of death from primary HNC.