High degrees of variability reported in Transcranial magnetic stimulation (TMS) assessment of corticospinal excitability (CSE) highlight the need to investigate its reliability as an assessment tool. This study investigated the effect of TMS test intensity on the amplitude, variability and test-retest reliability of motor evoked potentials (MEP). Twenty-five MEPs were recorded at 105%, 120%, 135%, 150% and 165% of resting motor threshold across three sessions in twelve participants. Repeated measures analysis of variance (RM-ANOVA), recruitment curve gradients and standardised z-value SDs were utilized to investigate the effect of TMS test intensity on MEP amplitude and variability. Test-retest reliability of MEP amplitude was assessed using RM-ANOVA and intraclass correlations (ICC). RM-ANOVA reported MEP amplitude significantly increased between 105-120% and 120-135% with non-significant increases thereafter. Recruitment curve gradients reduced as TMS test intensity increased. Standardised z-value SDs reported MEP amplitude variability reduced as TMS test intensity increased with the only significant reduction occurring between 120% and 135%. RM-ANOVA reported no significant effect of time on MEP amplitude, indicating agreement between sessions. ICCs indicated significant intra- and inter-session reliability for all TMS test intensities except for 105%. The only significant reduction in variability and final significant increase in MEP amplitude occurred between the same TMS test intensities. Reduced variability and increased reliability at higher TMS test intensities validate the use of higher TMS test intensities to assess changes in CSE in future research. To increase accuracy in capturing true changes in CSE we recommend assessing changes in CSE across a range of TMS test intensities.