Transcranial magnetic stimulation (TMS)-induced relaxation rate reflects intrinsic muscle contractile properties by interrupting the drive from the central nervous system during voluntary muscle contractions. To determine the appropriateness of knee-extensor muscle relaxation measurements induced by TMS, this study aimed to establish both the within- and between-session reliability before and after a fatiguing exercise bout. Eighteen participants (9 females, 9 males, age 25 ± 2 years, height 171 ± 9 cm, body mass 68.5 ± 13.5 kg) volunteered to participate in two identical sessions approximately 30 days apart. Maximal and submaximal neuromuscular evaluations were performed with TMS six times before (PRE) and at the end (POST) of a 2-min sustained maximal voluntary isometric contraction. Within- and between-session reliability of PRE values were assessed with intraclass correlation coefficient (ICC2,1, relative reliability), repeatability coefficient (absolute reliability), and coefficient of variation (variability). Test-retest reliability of post-exercise muscle relaxation rates was assessed with Bland-Altman plots. For both the absolute and normalized peak relaxation rates and time to peak relaxation, data demonstrated low variability (e.g. coefficient of variation ≤ 7.8%) and high reliability (e.g. ICC2,1 ≥ 0.963). Bland-Altman plots showed low systematic errors. These findings establish the reliability of TMS-induced muscle relaxation rates in unfatigued and fatigued knee-extensor muscles, showing that TMS is a useful technique that researchers can use when investigating changes in muscle relaxation rates both in unfatigued and fatigued knee-extensor muscles.