Collision-sport athletes, especially football players, are exposed to a higher number of repetitive head impacts. Little is known, however, regarding the effects of long-term exposure to repetitive head impacts on brain tissue structure and the locations (i.e. superficial or deep tissue structures) affected. On top of this, little is known about the effects of highly competitive athletics on brain tissue structure. We investigated this relationship, including the baseline effect of collegiate athletic participation, by mapping measures of microstructure to cortical and subcortical white matter parcels and major white matter tracts of varsity IU football players, cross-country runners, and non-athlete students using advanced microstructural mapping techniques and machine-learning. We tested a series of hypotheses on the differences between the subject groups. To do so, we implemented an innovative approach to statistical hypothesis testing combining advanced white matter tissue properties and machine-learning classifiers. We used cross-validation to select the best subject group model and best machine-learning classifier. Wide-spread differences in brain tissue microstructure across cortical, subcortical, and major white matter structures were documented. The tissue properties of the major white matter tracts were found to best predict collision-sport participation and sports participation in general, however, cortical and subcortical white matter parcels were also found to predict collision-sports participation. The biggest differences in brain tissue microstructure is between the athletes (football and cross country combined) and the non-athletes students. The rewards and risks of playing competitive sports at the highest collegiate level may account for the differences in brain microstructure. This was the first investigation into the effects of repetitive head impacts to use an open-source data processing platform brainlife.io. The data and code for these analyses are available via brainlife.io.