BackgroundExercise alters intestinal microbiota composition; however, it remains unclear whether these alterations induced by long-term endurance exercise can be transferred by fecal microbiota transplantation (FMT) to improve recipient exercise capacity.MethodsMice were divided into control (Group C), exercise (Group E), PBS transplantation (Group PT), and FMT (Group MT) groups. Group E underwent 14 weeks of treadmill training, and Groups PT and MT were transplanted with PBS or fecal microbiota from Group E, respectively. Subsequently, exhaustive exercise tests and 16S rRNA sequencing were performed, and blood glucose levels, glycogen reserves, and antioxidant indices were assessed.ResultsCompared with Group C, Group E had significantly higher hepatic/muscle glycogen levels and superoxide dismutase activity, and lower liver malondialdehyde content. After transplantation, Group MT showed significantly higher Firmicutes abundance, alpha diversity (Shannon and Simpson indices), and enrichment of eight beneficial genera (e.g., Bifidobacterium, Dorea) than Group PT, along with lower abundance of two harmful/conditional genera (e.g., Sutterella, Parabacteroides). Forty Kyoto Encyclopedia of Genes and Genomes metabolic pathways (e.g., starch and sucrose metabolism, flavone and flavonol biosynthesis) differed significantly between the PT and MT groups. The exhaustive exercise capacity of Group MT was significantly higher than that of Group PT (p < 0.05) but lower than that of Group E (p < 0.01). Correlation analysis showed that Sutterella abundance was negatively correlated with exercise capacity (r = −0.42, p < 0.05), whereas Dorea abundance was positively correlated (r = 0.48, p < 0.05).ConclusionExercise-induced increases in glycogen reserves and antioxidant capacity, associated with altered intestinal microbiota composition, can be partially transferred to recipient mice via FMT. The gut microbiota acts as a partial mediator of improvements in exercise capacity, rather than the sole driver.