The wide scope of our study is the set-up of a novel strategy called “Environmental Metabolic Footprinting” [1,2]. It aims to assess the environmental fate and impact of biocontrol products in soil, sediment and plants by tracking the evolution of their meta-metabolic profiles through the time. Thus, a broadband extraction covering a wide range of metabolites is required to allow for the best screening of those profiles. Hence, 5 extraction protocols (including 2 novel in-house-developed protocols) were tested on 3 environmental conditions (all applied on 2 soils of different types): control, spiked with the Glyphosate pesticide, spiked with Pelargonic acid formulation as a biocontrol product. However, determining the “optimal” extraction is not an easy task, as the experimental design is large, multi-factorial and complex (150 samples containing both soil biomolecules and pesticides components). Hence, an innovative LC-MS/HRMS-based untargeted approach was designed in order to evaluate the optimal protocol by using metabolomics computational tools (e.g. XCMS algorithm, OPLS-DA, Euclidean Distances, Heatmaps, van Krevelen, among others). Those tools provided easily-interpretable plots, comprehensive quantitative indices and chemical information based on ~500 metabolite annotations. They were used to assess 4 criteria defining the “optimal” conditions: 1) the ability of the extraction to cover a wide range of metabolites (qualitative), 2) its yield (quantitative), 3) its repeatability, and 4) its ability to allow for the discrimination between spiked and control samples. Therefore, the current presentation will introduce this innovative LC-MS/HRMS untargeted metabolomics-based approach as a powerful tool for methods development and will explain its different functions (“how it works”). The results of the large experiment mentioned above will be also highlighted. They allow promoting a novel broadband extraction (in-house-developed; based on solvent mixtures) as an optimal protocol for analyzing biocontrol products in soil samples. [1]: Patil et al. 2016. doi:10.1016/j.scitotenv.2016.05.071 [2]: Salvia et al. 2018. doi:10.1007/s11356-017-9600-6