BackgroundThis study determined the effect of video-verified collision characteristics on head impact magnitudes in male youth tackle football. MethodsParticipants (n = 23, age = 10.9 ± 0.3 years, height = 150.0 ± 8.3 cm, mass = 41.6 ± 8.4 kg) wore Triax Sim-G sensors throughout the fall 2019 season. Ten filmed games were used to identify nine different collision characteristics: mechanism, preparedness, head direction, struck versus striking activity, stance, play type, closing distance, penalty, and quarter. Random-effects general linear models and Cohen d effect sizes were used to examine differences in log-transformed peak linear (PLA; g) and rotational (PRA; rad/s2) accelerations across characteristics. The 10 games produced 533 total video-verified impacts and 23.2 ± 7.2 impacts per athlete. ResultsPLA (P range: 0.107 to 0.923) and PRA (P range: 0.057 to 0.768) did not differ across characteristics. Struck players (3370 rads/s2, 95% confidence interval [CI] = 2986 to 3808) had a small effect for higher PRA compared with striking players (3037 rads/s2, 95% CI = 2713 to 3404, d = 0.251), but negligible effect for simultaneous struck-striking players (3340 rad/s2, 95% CI = 2945 to 3792, d = 0.018). Fourth quarter impacts (3490 rads/s2, 95% CI = 3083 to 3951) had a small effect for higher PRA compared with first (2945 rads/s2, 95% CI = 2596 to 3337, d = 0.404), second (3196 rads/s2, 95% CI = 2832 to 3604, d = 0.219), and third quarters (3241 rads/s2, 95% CI = 2841 to 3699, d = 0.144). ConclusionYouth tackle football characteristics did not significantly affect head impact magnitudes during games. More research is needed to explore additional factors that could be modified for sport safety rather than mitigating impact mechanism.