To characterize the athlete microbiome, we have sequenced stool samples from marathon runners, ultramarathon runners, Olympic-caliber rowers, and sedentary controls (n=50) with each individual providing on average 7 samples (∼350 samples total). We have identified substantial alterations to the microbiome associated with athletic performance. In particular, we identified a bacteria genus, Veillonella, that increases in abundance after exercise. Veillonella’s primary source of energy is lactate - a metabolite that accumulates in the blood after exercise and is associated with fatigue. Through de novo assembly of ∼40 million contigs, we have developed a novel reference database from our athlete metagenomic/metatranscriptomic data sets - which consists of >2.5 million none redundant genes. Querying this database has yielded insights into potentially significant allelic variation, at the microbiome gene-level, involved in lactate related metabolic processes as well as multi’omic evidence of pathways that change between performance and recovery states. To validate this, we have isolated from our samples three species from this and introduced this cocktail of isolates to C57BL/6 mice prior to exhaustive exercise on a treadmill. Gavage of the Veillonella cocktail to mice four hours prior to exercise results in a significant decrease in serum lactate (lactate AUC of 40.31 mmol/L*hr relative to 66.07 mmol/L*hr in sham control mice; P < 0.05). Follow-up work currently being conducted involves assessing the role of Veillonella’s primary metabolic end product, the short chain fatty acid propionate, on limiting exercise-induced inflammation, and the effects of these components on recovery from exercise. Taken together, our study encompasses both high-resolution metagenomic analysis and experimental validation to understand the role that the microbiome naturally plays in the performance of elite athletes. Disclosure A. Kostic: None.