The purpose of this study was to analyze the impact of tumor volume (TV) measurements as prognosticator for recurrence-free survival (RFS) and overall survival (OS) from data of head and neck cancer (HNC) registries. TV measurements were performed in pre-treatment computed tomography (CT) or magnetic resonance images (MRI) of 392 unselected HNC patients. TV measurements were feasible in 275 patients (70 %). Median CT TV and MRI TV were 11.43 and 10.4 cm(3), respectively. The CT TV was significantly different only between T1 and T4. CT TV was significantly different only between T1 and T4 (p = 0.041). MRI TV was significantly different between T1 and T4 (p = 0.003) as well as between T2 and T4 (p = 0.002). Median follow-up was 26.1 months. Median RFS was 80.7 months. Median OS was 66.5 months. On univariate analysis, significant prognostic factors for decreased RFS were advanced T stage (p = 0.010); M1 (p = 0.001) and an MRI TV > 10.4 cm(3) (p = 0.001). Significant prognostic factors for a decreased OS were advanced T stage (p = 0.001), N+ (p = 0 006), M+ (p < 0.001), tumor recurrence (p < 0.001), CT TV (p = 0.005), and MRI TV (p = 0.012). On multivariate analysis for RFS, MRI TV was the best independent prognosticator (p = 0.003). On multivariate analysis for OS, T stage (p = 0.006) was a better prognosticator than CT or MRI TV. Using CT and MRI data sets of an unselected series of HNC patients in a cancer registry, TV measurements were not feasible in all patients. MRT TV was a powerful prognosticator for RFS.