Introduction 18 Fluorine-fluorodeoxyglucose positron emission tomography/computed tomography is not yet accepted as a standard pretreatment evaluation of thymic epithelial neoplasm (TEN). Statistical correlation between standardized uptake value of tumor/mediastinum ratio and patients’ WHO risk class has been reported. PET metabolic tumor volume (MTV) and total glycolytic volume (TGV) have been reported as additional prognostic imaging biomarkers in several human tumors. Purpose of study was to establish whether MTV and TGV add prognostic information in TEN. Materials and methods A retrospective dynamic cohort study of prospectively collected data (2006–2012) on 23 consecutive patients with pathologically proven TEN (no thymic carcinoma) was conducted. All patients underwent chest CT, and PET for staging. SUV T/M ratio, semi-quantitative and volumetric analyses of TEN were calculated. Patients were categorized according to WHO classification (low-risk and high-risk thymomas). Statistical analysis was performed with bootstrap method. Multi-collinearity was established using Pearson correlation coefficient. Cut-off point for TGV was compared using Mantel Cox log rank test. Results SUV T/M ratio, MTV, and TGV correlate with low- and high-risk TEN. However, the statistical correlation between TGV and WHO classification ( ρ  = 0.897) was higher than SUV T/M ratio ( ρ  = 0.873). Since sample distributions were not uniformly smooth, only one cut-off value was identified: a TGV of 383 served as a cut-off value between low-risk and high-risk TEN. Conclusion TGV is a PET reproducible imaging marker in patients with TEN, provides prognostic information, and could be useful in pretreatment stratification of patients. Nevertheless, it needs validation in larger cohort studies.