Gentiana lutea rhizomes are known for their bitter tasting properties conferred by its unique biochemical content. They are currently of interest in phytotherapy, animal nutrition, food processing, cosmetic applications and agroecology. In this study, a NIRS, fluorescence and HS-SPME-GCMS dataset of 55 rhizomes from four different French mountains (Alpes, Jura, Massif Central and Pyrénées) was collected with the aim of assessing the variability of Gentiana lutea composition at different scales. The feasibility of data fusion strategies was demonstrated to be effective in distinguishing the geographical origin of Gentiana lutea roots over a wide area. The results suggest that data fusion methods have the potential to be more effective in the quality of separation of studied sites of Gentiana lutea roots than individual decisions obtained from individual analytical tools. However, to guarantee the geographical origin of Gentiana lutea roots within a single massif using these techniques, environmental factors must be considered.