In threshold-tracking TMS, stimulation intensity is adjusted online to maintain a predefined target MEP response, thus allowing monitoring and controlling for the naturally occurring fluctuations in corticospinal excitability as well as technical errors. Although it is emerging as a useful clinical tool, little is known about its reliability and comparability with conventional techniques.To evaluate the reproducibility of the threshold-tracking (constant response) TMS technique for short interval intracortical inhibition (SICI) and compare it with standard (constant stimulus) measurements.SICI was measured in 12 healthy volunteers (6 males, 31±7.6years) from the right FDI muscle using an automated stimulation protocol incorporating (i) conventional, fixed-intensity (A-SICI) and (ii) threshold tracking (T-SICI) methods at conditioning stimuli (CS) of 50–80% resting motor threshold (RMT) and interstimulus interval of 2.5ms. RMT was defined as a stimulation intensity required to evoke a response of 0.2mV and was determined by threshold tracking. Measurements were repeated on the same day and at least a week later by a single operator.No significant differences in mean RMT, A-SICI and T-SICI were observed between sessions. RMT and TS, measured by threshold-tracking technique, showed excellent intra- and adequate interday reproducibility. At 50% RMT CS both types of SICI had poor reproducibility. At all other CS intensities, T-SICI had excellent intraday (intraclass correlation coefficient, ICC 0.81–0.92) and adequate-to-excellent interday (ICC 0.61–0.88) reproducibility, while A-SICI showed poor intraday and poor-to-adequate interday reproducibility (ICC 0.29–0.37 and 0.37–0.51, respectively).SICI measured by threshold tracking shows good intra- and interday reproducibility. Automation of the procedure and allowance for naturally occurring fluctuations in cortical excitability by threshold tracking may increase the reproducibility of SICI.