Despite their known risks, herbicides are still essential for agriculture. Thus, natural herbicides are increasingly recommended to replace synthetic ones. They are supposed less harmful on human health and ecosystem. However, there are still limitations in studying the environmental fate of several of them. The main reason is the lack of methods dedicated for this type of herbicides. In fact, they usually consist of complex mixtures of substances, of which several are unknown. Hence, in the framework of the Environmental Metabolic Footprinting (EMF) approach, the current work presents a method dedicated to analyze the volatile residues of herbicides applied on soil. The aim is to track the evolution of the volatile compounds issued from the herbicide by an untargeted metabolomics-based kinetics, in order to determine the “resilience time” of the gaseous phase above the soil. The approach aims to explore the environmental fate of these herbicides by combining it to other methods of the EMF. Moreover, it can estimate the exposure of farmers, insects and plants to potential toxic volatile substances. The method is based on Headspace-Solid Phase Micro Extraction-Gas Chromatography-Quadrupole Mass Spectrometry (HS-SPME-GC-Q MS). The HS-SPME provides a non-destructive extraction. This allows reducing the number of samples. The GC separation technique provides high reproducible analysis for volatile compounds. In addition, it allows the calculation of the Retention Index (RI) as a tool for the molecular characterization. On the other hand, despite the low resolution of the Quadrupole mass analyzer, the Electronic Impact provides reproducible MS fragmentation patterns used for spectral library search and putative identification of unknown compounds. The method was experimented for a pilot study. It was applied on a natural herbicide: the extract of Myrica gale, containing the Myrigalone A active substance. The setup of the analytical tool was performed by optimizing the system parameters. Then, a 38 days kinetics study was applied on control and spiked soil samples. Results proved the reliability of the method and demonstrated the robustness of the system and low matrix effect. Thus, new high scale experiments are planned. The integration of other natural and synthetic herbicides will be also considered.