Emerging evidence highlights the role of gut microbiome in mental health disorders, including depression, raising the question whether the action of antidepressants could be mediated, at least in part, via the microbiome-gut-brain axis. To explore this, we subjected a St. John's wort extract (STW 3-VI), clinically proven to be effective in mild to moderate depression, to a model of the upper and lower intestinal tract, including static in vitro predigestion followed by ex vivo incubation with human microbiota samples. To cover the interindividual diversity of gut microbiome composition, fecal samples from ten healthy volunteers were used. Although unchanged levels of most annotated compounds were observed during simulated upper intestinal tract digestion, incubation with fecal microbiota led to a significant change of the chemical profile of the extract. While hyperforins remained stable, flavonoids and hypericins were rapidly biotransformed, suggesting that they may act as prodrugs. Several metabolites were formed, many of which are known to be involved in gut-brain communication. Differential abundance analysis revealed significant changes in microbiome composition, particularly for taxa known to be potentially associated with depression. Among others, the Firmicutes / Bacteroidetes ratio, known to be lowered in depressive patients, was increased. Functional profiling revealed modulation of pathways involved in gut-brain communication, such as tyrosine and tryptophan metabolism. These bidirectional interactions suggest for the first time the gut microbiome as a potential mediator of the pharmacological effects of St. John's wort extracts via the microbiome-gut-brain axis.