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. Please refer to Ghosson, H. et al. Anal. Chim. Acta (2020), 1134:58-74. doi:10.1016/j.aca.2020.08.016.