Instrumented mouthguards are frequently used to study head impacts in sports. However, we hypothesise that repeated usage of the mouthguard may compromise the fit on the teeth and result in lower data quality. Sensor acceleration events were recorded from youth American football players. Events verifiable on video were included and subjected to the mouthguard manufacturer’s classification algorithm. For each event, the device provides a quality class (high, moderate, or low), which we examined for the various data cleaning methods, and in relation to the duration of mouthguard usage. Out of all events (n=9,375), the proportions of high-, moderate-, or low-quality events were 8%, 28%, and 64%, respectively. Among those, 184 events were determined as true positives by video review and algorithm, with high-, moderate- and low-quality proportions of 81%, 14%, and 5%. There was no correlation between mouthguard usage duration and signal quality for videoverified events, however, there was a positive correlation when including all events (r=0.170, p=0.031), which is contrary to our hypothesis. Whether the kinematics from low-quality events are valid is unknown, and whether these should be included in analyses is questionable. Ongoing analysis of data from practices may provide further clarification on signal quality and mouthguard wear-and-tear.