Proteomic analysis of plasma offers tremendous insight into tissue to tissue communication and signalling in complex scenarios such as exercise. Inspired by the growing appreciation that such cross talk might be partially mediated by proteins packaged in extracellular vesicles, we have carried out a deep analysis of the exercise-induced extracellular vesicle proteome. Using a series of centrifugation and PBS wash steps, we isolated extracellular vesicles from the arterial plasma of 11 human participants carrying out a 1hr bout of cycling. All samples were analysed by nano-ultra high-pressure liquid chromatography coupled to tandem MS on a Q-Exactive MS operated in DDA. In addition to experimental samples taken at rest, immediately and 4 hours after exercise, we also pooled exercise and baseline samples and fractionated via neutral pH HPLC on an in-house micro packed BEH C18 column to derive a deep reference proteome. Raw data were searched against the human uniprot database in Maxquant, utilising the match between runs function to map and transfer identifications from the reference data set to the experimental samples. We identified a total of 5359 proteins at an FDR of 0.01, with the large majority of this coverage derived from the fractionated reference proteome. Matching between runs resulted in a 105% gain of identifications in experimental samples, facilitating a quantitative LFQ comparison of 1159 proteins. Analysis in Perseus revealed 325 proteins to be differentially regulated by exercise with a notable upregulation of several classes of proteins that compose the canonical 40-150nm exosome. Proteins associated with the formation (ALIX), trafficking (RAB GTPases, Annexins) and target adhesion (Integrins, tetraspanins) of exosomes were all upregulated by exercise. Pathway analysis revealed significant enrichments in several biological processes and signalling pathways. This raises the possibility that exercise mediates many of its affects via the release of key signalling molecules packaged in exosomes.