The aim of this study was to determine the applicability and reliability of a transcranial magnetic stimulation twitch interpolation technique for measuring voluntary activation of a lower limb muscle group. Cortical voluntary activation of the knee extensors was determined in nine healthy men on two separate visits by measuring superimposed twitch torques evoked by transcranial magnetic stimulation during isometric knee extensions of varying intensity. Superimposed twitch amplitude decreased linearly with increasing voluntary torque between 50 and 100% of mean maximal torque, allowing estimation of resting twitch amplitude and subsequent calculation of voluntary activation. There were no systematic differences for maximal voluntary activation within day (mean ± S.D. 90.9 ± 6.2 versus 90.7 ± 5.9%; P = 0.98) or between days (90.8 ± 6.0 versus 91.2 ± 5.7%; P = 0.92). Systematic bias and random error components of the 95% limits of agreement were 0.23 and 9.3% within day versus −0.38 and 7.5% between days. Voluntary activation was also determined immediately after a 2 min maximal voluntary isometric contraction; in four of these subjects, voluntary activation was determined 30 min after the sustained contraction. Immediately after the sustained isometric contraction, maximal voluntary activation was reduced from 91.2 ± 5.7 to 74.2 ± 12.0% (P <0.001), indicating supraspinal fatigue. After 30 min, voluntary activation had recovered to 85.4 ± 8.8% (P = 0.39 versus baseline). These results demonstrate that transcranial magnetic stimulation enables reliable measurement of maximal voluntary activation and assessment of supraspinal fatigue of the knee extensors.