The efficacy of motor imagery (MI) practice to facilitate muscle stretching remains controversial and the underlying neurophysiological mechanisms unexplored. We evaluated the effects of MI practice during a sit-and-reach task. Healthy participants were randomly assigned to a MI practice (n = 15) or Control (n = 15) group and completed 2 blocks of 5 sit-and-reach trials. During the first block (B1), participants performed 5 maximal stretching trials of 10s. During the second block (B2), trials were divided into two consecutive parts: i) reproducing the maximum performance of B1 (10s, B2 part 1), and ii) attempting to outperform the maximum performance of B1 (10s, B2 part 2). Participants performed kinesthetic MI of hamstring stretching during B2 trials in the MI practice group. We recorded electromyography from the hamstring and rectus femoris of the dominant leg. We also processed skin conductance as an index of sympathetic activity. We measured greater performance improvements from B1 to B2 part 2 in the MI practice group compared to Control (p < 0.05). Participants in the MI practice group exhibited reduced hamstring activation during both B2 part 1 (p < 0.001) and B2 part 2 (p < 0.001) compared to Control. Skin conductance revealed higher sympathetic activation during B2 part 2 compared to both B1 and B2 part 1 in the two groups. Thus, performing MI during actual movement is likely to improve stretching performance through reduced muscle activation. Such improvement may be grounded in a cortical gain over spinal reflexes.