The objective of this study was to investigate the value of metabolic tumour volume (MTV) assessed with (18)F-FDG PET/CT in predicting event-free survival (EFS) and overall survival (OS) in patients with head and neck squamous cell carcinoma (HNSCC), and particularly to compare it with more conventional parameters such as maximum standardized uptake value (SUVmax). Patients referred to our department for (18)F-FDG PET/CT for staging of HNSCC were prospectively included between February 2009 and March 2011. Each patient was scanned using a Philips Gemini PET/CT system at 1 h after injection. The MTV was calculated semiautomatically for the primary site using methods based on SUV with various thresholds: 3-D contour around voxels equal to or greater than 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5 and 7.0 times SUV, or more than 30%, 40% and 50% of SUVmax. ROC analysis was used to test the statistical significance of the differences among the calculated MTVs. EFS and OS were determined using the Kaplan-Meier method and compared with MTV in univariate and multivariate analyses, including the usual prognostic factors: age, sex, primary site, treatment, SCC histologic grade, AJCC stage, TNM classification, tumour SUVmax and SUVpeak. The study included 80 consecutive patients (70 men, 10 women; mean age 62.4 ± 9.0 years). ROC analysis revealed that pretreatment MTV using a threshold of 5.0 times SUV (MTV5.0) was the best parameter to predict recurrence and death after treatment. In univariate analysis, MTV5.0 >4.9 ml was predictive of poor EFS (p < 0.0001) and poor OS (p < 0.0001). In multivariate, MTV5.0 persisted as an independent predictive factor for EFS (p = 0.011) and OS (p = 0.010), while SUVmax became nonsignificant (p = 0.277 for EFS, p = 0.975 for OS). Our results suggest that MTV measured by (18)F-FDG PET/CT has independent prognostic value of in patients with HNSCC, stronger than SUVmax.