BACKGROUND: Animal and human studies have revealed reciprocal association between exercise and gut-brain axis. However, the clinical evidence from randomized controlled trials (RCT) are still limited to directly assess the effects of aerobic exercise on gut microbiota. To fill this gap, we conducted this 12-week RCT in both groups of adolescents with and without sub-threshold mood symptoms. METHODS: A total of 224 adolescents were randomized to the aerobic exercise intervention or psychoeducation-controlled arm. 49 adolescents with subthreshold symptoms and 142 clinically-well adolescents provided the sample for microbiota assessed by metagenomic sequencing. Aerobic exercise of running at the moderate-intensity for 30 min per day, 5 days a week, were conducted for 12 weeks. RESULTS: Adolescents with subthreshold symptoms had significantly lower beta diversity than clinically-well adolescents in both the exercise intervention and psychoeducation-controlled arms (p<0.05). After intervention, no difference in gut microbiota diversity, phylum, genus, species level abundancies or gut microbial functions were found in both of the symptomatic or non-symptomatic groups. Metagenome-wide association study analysis showed no significant difference in metagenomic linkage groups. LIMITATIONS: The sample size is relatively small. The exercise intensity we employed may be insufficient to result in observable effects on intestinal microbiota. CONCLUSION: We conclude that a 12-week moderate-intensity aerobic exercise intervention showed no significant beneficial effect on the gut microbiota in clinically-well adolescents as well as in adolescents with subthreshold symptoms. The beta diversity of gut microbiota in adolescents with subthreshold mood syndromes may be impaired when compared with clinically-well adolescents.