Electromechanical delay (EMD) is the latency between the onset of electrical activity and force in a muscle. While series elastic component is thought to be the major contributor to EMD, the relative contribution of other factors (propagation of the action potential and excitation-contraction coupling) is still undefined. The first aim was to assess the reliability of an electromyographic (EMG) and mechanomyographic (MMG) approach in partitioning EMD. Secondly, the effect of fatigue on the several EMD components was evaluated. Tetanic stimulations (5s at 50Hz) were delivered to the medial gastrocnemius of 12 subjects, before and after a fatiguing stimulation (120s at 35Hz). On a different day, the study was repeated to assess reliability. During contractions, EMG, MMG and force were recorded. EMD and delays between EMG and MMG (Dt EMG-MMG, EMD muscle component) and MMG and force (Dt MMG-F, EMD tendon component) were calculated. Intraclass correlation coefficient (ICC) was 0.892, 0.915 and 0.917 for EMD, Dt EMG-MMG and Dt MMG-F, respectively. With fatigue, peak force drop (-26%) was accompanied by significant EMD, Dt EMG-MMG and Dt MMG-F elongation from 25.58±2.62 ms, 11.44±0.64 ms and 13.84±2.96 ms to 31.57±4.18 ms (+23%), 12.75±0.66 ms (+12%) and 18.98±4.44 ms (+39%), respectively. ICC showed that EMG-MMG approach is a reliable tool to partition EMD. Although fatigue had a stronger effect on the tendon component of EMD, also muscle component was significantly affected.