Purpose:
This exploratory study compared changes in squat strength and kendo men-strike biomechanics between velocity-based training (VBT) and percentage-based training (PBT) conditions.

Methods:
Twelve university kendo athletes completed a 7-week back-squat intervention under VBT (n = 7) or PBT (n = 5). Allocation was based on academic year rather than randomisation. Before and after training, squat one-repetition maximum (1RM), men-strike motion time, push-off-phase rate of force development over 0–100 ms (RFD0–100 ms), and the temporal overlap of high-extension-torque phases among the hip, knee, and ankle joints were assessed.

Results:
Absolute and relative squat 1RM increased in both conditions. Between-condition differences in change were 14.1 kg for absolute 1RM (95% CI [4.5, 23.8]) and 0.309 kg · kg^–1 for relative 1RM (95% CI [0.136, 0.482]), favouring VBT. The between-condition difference for RFD0–100 ms also favoured VBT (27.62 N · kg^–1 · s^–1; 95% CI [–11.75, 66.98]), but the CI included zero. Differences were small and uncertain for men-strike motion time (–0.025 s; 95% CI [–0.142, 0.092]) and the overlap index (1.6 percentage points; 95% CI [–4.8, 8.1]).

Conclusions:
Between-condition differences were clearest for squat strength, whereas evidence for kendo-specific biomechanical changes was limited. Because academic year was confounded with intervention condition, and the interventions differed in load progression and autoregulation, the findings cannot be attributed to velocity feedback alone. Evidence for transfer to faster men-strike performance remains inconclusive.