This study compared the effects of variable-resistance compound training versus constant-resistance compound training on lower-limb explosive power in judo athletes, aiming to identify an effective and safe training method. Methods: Sixteen judo athletes were randomized into VRT (n = 8) or RT (n = 8) groups for a 6-week, twice-weekly intervention. Outcomes included rate of force development (RFD), counter movement jump (CMJ), squat jump (SJ), eccentric utilization ratio (EUR), reactive strength index (RSI), squat 1RM, muscle architecture, and Special Judo Fitness Test (SJFT). After 6 weeks of training intervention, the time × group interaction effects were significant between the variable-resistance compound training group and the constant-resistance compound training group in the following parameters: CMJ (p < 0.01, η² = 0.605), SJ (p < 0.01, η² = 0.391), EUR (p < 0.01, η² = 0.308), RSI (p < 0.01, η² = 0.306), RFD (p < 0.01, η² = 0.401), squat 1RM (p < 0.01, η² = 0.328), SJFT index (p < 0.01, η² = 0.537), femoral rectus feather angle (p < 0.05, η² = 0.380), femoral rectus thickness (p < 0.05, η² = 0.288), femoral rectus cross-sectional area (p < 0.01, η² = 0.868), gastrocnemius feather angle (p < 0.05, η² = 0.274), and gastrocnemius thickness (p < 0.05, η² = 0.390). No significant group effects were observed for any of the parameters (p > 0.05). Conclusion: Both variable-resistance training (VRT) and constant-resistance training (CRT) are effective in enhancing lower-body power in judo athletes; both training methods can be regarded as effective options for developing lower-body power in judo athletes, although VRT may offer a slight advantage in specific performance domains.