We investigated the effects of the initial length of the muscle-tendon unit (MTU) and muscle pre-activation on muscle-tendon interactions during plantarflexion performed at maximal velocity. Ultrasound images of gastrocnemius medialis were obtained on 11 participants in three conditions: (a) active plantarflexion performed at maximal velocity from three increasingly stretched positions (10°, 20°, and 30° dorsiflexion), (b) passive plantarflexion induced by a quick release of the ankle joint from the same three positions, and (c) pre-activation, which consisted of a maximal isometric contraction of the plantarflexors at 10° of dorsiflexion followed by a quick release of ankle joint. During the active condition at maximal velocity, initial MTU stretch positively influenced ankle joint velocity (+15.3%) and tendinous tissues shortening velocity (+37.6%) but not the shortening velocity peak value reached by muscle fascicle. The muscle fascicle was shortened during the passive condition; however, its shortening velocity never exceeded peak velocity measured in the active condition. Muscle pre-activation resulted in a considerable increase in ankle joint (+114.7%) and tendinous tissues velocities (+239.1%), although we observed a decrease in muscle fascicle shortening velocity. During active plantarflexion at maximal velocity, initial MTU length positively influences ankle joint velocity by increasing the contribution of tendinous tissues. Although greater initial stretch of the plantarflexors (ie, 30° dorsiflexion) increased the passive velocity of the fascicle during initial movement, its peak velocity was not affected. As muscle pre-activation prevented reaching the maximal muscle fascicle shortening velocity, this condition should be used to characterize tendinous tissues rather than muscle contractile properties.