Sports-related concussions were once thought to present only transient symptoms, but recent research has shown the potential for long-term neurological impairments. Previous research has instrumented helmeted athletes with sensors to measure head kinematics and concussive biomechanics on-field. However, expanding studies beyond helmeted players will help us to understand sex-specific differences and unhelmeted loading conditions and their relationship to impact exposure and tolerance. Athletes on the men’s and women’s club collegiate rugby teams at Virginia Tech were custom fit with instrumented mouthguards and monitored over a year to quantify sub-concussive and concussive biomechanics in unhelmeted athletes. The objective of this study was to quantify head kinematics sustained from inertially driven (body impacts) and direct head impacts in collegiate rugby players. The majority of the impacts collected were low in magnitude, and head impacts were associated with higher PLAs than body impacts (∆ = 2.5 g, p < 0.005). Head impacts had a median PLA of 15.2 g and a median ∆RV of 6.7 rad/s;​ body impacts had a median PLA of 13.6 g and a median ∆RV of 6.6 rad/s. For each impact type, men and women sustained similar peak linear accelerations (head impact:​ ∆ = 1.9 g, p = 0.219;​ body impact:​ ∆ = 0.5 g, p = 0.280). Men sustained higher changes in rotational velocities for head impacts (∆ = 1.4 rad/s, p = 0.003) and longer impact durations for both head (∆ = 2.6 ms, p<0.001) and body (∆ = 2.8 ms, p = 0.023) impacts. These findings can be used to advise best practices for in-mouth sensors and better understand the mechanisms of concussion in unhelmeted sports.