Purpose: Our study aimed to explore the optimal timing as well as the most appropriate prognostic parameter of ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG PET) during chemoradiotherapy (CRT) for an early prediction of outcome for patients with head and neck squamous cell carcinoma (HNSCC). Methods: Serial PET data (before and three times during CRT) of 37 patients with advanced stage HNSCC, receiving combined CRT between 2005 and 2009, were evaluated. The maximum standardized uptake value (SUV_max), the average SUV (SUV_mean) and the gross tumour volume determined by FDG PET (GTV PET), based on a source to background algorithm, were analysed. Stratified actuarial analysis was performed for overall survival (OS), disease-free survival (DFS) and locoregional control (LRC). The median follow-up time was 26 months (range 8-50). Results: For all patients, OS was 51%, DFS 44% and LRC 55% after 2 years. The 2-year OS (88%) and 2-year LRC (88%) were higher for patients whose SUV_max of the primary tumour decreased 50% or more from the beginning (0 Gy) to week 1 or 2 (10 or 20 Gy) of CRT (ΔSUV_max10/20≥50%) than for patients with ΔSUV _max20<50% (2-year OS=38%; p=0.02; 2-year LRC 40%; p=0.06). A pretreatment GTV PET below the median of 10.2 ml predicted a better 2-year OS (34% for GTV PET≥10.2 ml vs 83% for GTV PET<10.2 ml; p=0.02). Conclusion: The decrease of SUV_max from before (0 Gy) to week 1 or 2 (10 or 20 Gy) of CRT is a potential prognostic marker for patients with HNSCC. Because GTV PET depends on the applied method of analysis, we suggest the use of SUV _max, especially ΔSUV_max10/20, for an early estimation of therapy outcome. Confirmatory studies are warranted.