This work introduces a novel online Headspace-Solid Phase Microextraction-Gas Chromatography-Mass Spectrometry-based untargeted metabolomics approach, suggested as an alternative tool to study the environmental fate of volatile xenometabolites in emerging complex biopesticides, e.g. the Myrica gale methanolic extract herbicide containing several unknown metabolites. A “living” microcosm sample was designed for non-destructive analysis by a 35-min HS-SPME automated extraction and a 36-min GC-MS run. A 38-day kinetics study was then applied on two groups of soil samples: control and spiked. Statistical tools were used for the comparative kinetics study. The Principal Component Analysis revealed and explained the evolution and the dissipation of the herbicide volatile xenometabolome over time. The time-series Heatmap and Multivariate Empirical Bayes Analysis of Variance allowed the prioritization of 101 relevant compounds including 22 degradation by-products. Out of them, 96 xenometabolites were putatively identified. They included 63 compounds that are identified as herbicide components for the first time. The Orthogonal Projections to Latent Structures Discriminant Analysis and its Cross-Validation test were used to assess the total dissipation of the herbicide volatile residues and method detection limit. The reproducibility of the method was also assessed. The highest inter-samples (n = 3) Peak Area RSD was 7.75 %. The highest inter-samples (n = 3) and inter-days (n = 8) Retention Time SD were 0.43 sec and 3.44 sec, respectively. The work presents a green, non-laborious and high-throughput approach. It required a small number of environmental samples (6 microcosms) that were analyzed 8 times and were not destroyed during the study.