We examined the effects of five loading conditions (0%, 20%, 40%, 60%, and 80% of body-mass [BM]), on resisted sprint performance and kinematics in male rugby players over different distances. Ten players from the Brazilian National Team (20.1±3.3 years; 88.7±18.8 kg; 178.3±6.2 cm) performed 20-m sprints under the five loading conditions. Sprint times in 5-, 10- and 20-m were recorded. Stride length (SL), and hip, knee and ankle angles were measured using an eight-sensor motion analysis system. The kinematic parameters were calculated over the different distances. Heavier loads led to significantly greater velocity loss (P < 0.001-0.05). Significant reductions in SL were also observed when comparing 0% BM and all resisted sprints in all assessed distances (P < 0.001-0.05, Effect Size, [ES]: 1.35-4.99). Very-heavy (80% BM) sled load provoked significantly greater decreases in SL than the rest of loading conditions (P < 0.01-0.05). Important kinematic alterations were observed for all loading conditions and sprint distances when compared to 0%BM (ES: 0.76-1.79, for hip-angle; 0.20-1.40, for knee-angle; and 0.73-1.88, for ankle-angle). Moreover, 80% BM induced significantly higher hip flexion, lower knee flexion and higher ankle dorsiflexion than 20% BM condition at 5-10- and 10-20-m distances (P < 0.05). Lighter sled loads (< 40% BM) seem to be more adequate to improve speed ability without provoking drastic changes in unloaded sprinting technique, whereas heavier loads might be more suitable for optimizing horizontal force production and thus, acceleration performance.