PURPOSE: This study assessed the test-retest reliability of TMS motor evoked potentials (MEPs) and silent periods at early, middle, and late phases of the rising time-torque curve during explosive voluntary contractions. We also investigated how the number of consecutively averaged measurements influenced reliability. METHODS: On two separate occasions 3-7 days apart, 14 adults performed several isometric explosive (1-s) contractions of the knee extensors, some of which were superimposed with TMS to elicit MEPs in the superficial quadriceps. Of those with TMS, stimulation was timed to elicit MEPs at either 45 (early), 115 (middle), or 190 ms (late) following contraction onset (16 with-TMS contractions per time condition). TMS was also superimposed at the plateau of 15 separate MVCs. Test-retest intraclass correlation coefficient (ICC) and coefficient of variation (CV) were calculated for MEPs and silent periods consecutively averaged over 3 to 15 separate contractions. RESULTS: No one condition/phase was more reliable than another. For MEP amplitude, in all conditions except the explosive late phase, ICCs generally increased, and CV decreased, with an increase in the number of averaged contractions, and were >
0.50 ICC and <
15% CV within seven contractions. For silent period, ICCs and CVs were unaffected by the number of consecutively averaged contractions and remained >
0.50 ICC and <
10% CV. CONCLUSION: Test-retest reliability of TMS responses is comparable between phases of explosive contraction and at the plateau of MVC. To maximise reliability of MEPs during explosive contractions or MVCs, we recommend future studies average data across more than the 3-5 contractions typically reported in the literature investigating MEPs at MVC plateau.