Instrumented mouthguards measure on-field sports head impact and concussion biomechanics. While an increasing number of studies have applied mouthguard sensors, potential kinematic error associated with on-field mouthguard coupling variations have not been quantified. In addition, past research has been heavily biased in study population, lacking sex and sport diversity. Studies of concussion mechanisms have also primarily concentrated on single severe impacts with limited investigation of repeated impact exposure mechanisms. This research aims to assess and apply instrumented mouthguards in on-field settings to investigate head impact biomechanics and their injury influence in sports. We deployed mouthguard sensors to university women’s rugby and men’s hockey athletes during competitive seasons. Coupling categories were defined using on-field proximity sensor data before comparing signal characteristics between categories. Additionally, impactsfrom women’s rugby and men’s hockey were characterized using head kinematics- and brain deformation-based injury criteria. The cumulative effect of head impacts on concussion risk was investigated by modifying and applying a time-weighted exposure injury criterion to concussion and non-concussion datasets. Results showed that poorly coupled mouthguards led to higher angular kinematics and high-frequency noise. In our impact biomechanics characterization, women’s rugby impacts showed lower angular kinematics. Our modified time-weighted exposure metric showed increased sensitivity in differentiating concussion and non-concussion days compared to the same-day impact frequency, timing, and magnitudes separately. From our findings, we demonstrate the potential for using proximity sensors in instrumented mouthguards to assess on-field head impact coupling and data quality. Characterization of women’s rugby and men’s hockey head impacts and comparison with existing literature revealed that impact kinematics may vary by sensor type, age, sport, and sex. These differences must be accounted for when developing injury mitigation techniques, particularly in women’s sports. Finally, considering head impact history may allow for increased sensitivity in predicting concussions. In summary, this thesis contributes novel insights into on-field instrumented mouthguard data quality, potential sport- and sex-based differences in head impact exposure, and the role of repeated impact exposure in concussion risk. The findings from this work may be applied in future research for more rigorous sensor-based investigation of sports head impacts and concussions.