The purpose of this study was to clarify the influence of muscle-tendon complex stretch on electromechanical delay (EMD) in terms of the extent of tendon slack in the human medial gastrocnemius (MG). EMD and MG tendon length were measured at each of five ankle joint angles (-30, -20, -10, 0, and 5 degrees : positive values for dorsiflexion) using percutaneous electrical stimulation and ultrasonography, respectively. The extent of MG tendon slack was calculated as MG tendon length shortening, standardized with MG tendon slack length obtained at the joint angle (-16 degrees +/- 5 degrees ) where the passive ankle joint torque was zero. EMD at -30 degrees (19.2 +/-2.2 ms) and -20 degrees (17.2 +/- 1.3 ms) was significantly greater than that at -10 degrees (16.0 +/-2.3 ms), 0 degrees (15.0 +/-1.4 ms), and 5 degrees (14.8 +/-1.4 ms), and at 0 and 5 degrees, respectively. The relative EMD, normalized with the maximal EMD for each subject, decreased dependent on the extent of decrease in MG tendon slack. There were no significant differences in EMD among the joint angles (-10, 0, and 5 degrees ) where MG tendon slack was taken up. These results suggest that the extent of tendon slack is an important factor for determining EMD.