It is well known that when pairs of transcranial magnetic stimulation (TMS) pulses are applied to the motor cortex, the corresponding motor evoked potential (MEP) responses are altered compared to those of single pulses. At appropriate interstimulus intervals (ISIs) paired pulses are facilitatory, and this is believed to result from summation of I-waves by the secondary TM stimulus (Ziemann et al. 1998). If this is the case it should be possible to further enhance MEP responses by delivering additional stimuli at I-wave periodicity in order to target later I-waves. Therefore we have compared double pulse TMS responses with those of triple pulse TMS (tTMS) at varying ISIs. In seven healthy subjects (30-45 years) TMS was delivered to the optimal scalp position for MEP responses from the dominant first dorsal interosseus muscle using a 9cm diameter figure-8 coil held tangentially to the skull. Resting (RMT) and active motor thresholds (AMT) to single pulse TMS (Magstim 200) were determined for each subject and subsequent double and tTMS (3 x Magstim 200 units connected to a single coil) were delivered in the resting condition at an intensity equivalent to AMT. Double pulses (condition A) were delivered at an ISI of 1.5ms, whilst the 3 tTMS conditions comprised ISIs between stimuli 1-2 and 2-3 of 1.5/1.5ms (B); 1.5/2.0ms (C) and 1.5/3.0ms (D), respectively. Ten tTMS stimuli for each condition were delivered pseudo-randomly with 5 s intervals between each stimulation. Repeated measures ANOVA was used to determine the effects of condition and t-tests were performed post-hoc using Bonferroni correction as appropriate. Values are mean ± SEM and the overall level of significance is p < 0.05. No MEPs were elicited using single pulse TMS at the stimulus intensity used (since the muscle was non-active and RMTs were approximately 10% higher than AMTs). The double pulse condition resulted in discernable MEPs in all subjects (condition A = 0.81 ± 0.19 mV). Triple pulses at the same intensity resulted in larger MEPs and these differences were significant for conditions B (2.33 ± 0.32 mV; p < 0.002) and D (1.48 ± 0.36 mV; p 0.05). Comparing triple pulses, condition B with the shortest S2-S3 ISI elicited significantly larger responses than the conditions with longer S2-S3 intervals (condition C; p < 0.01 and condition D; p < 0.05). Conditions C and D were not significantly different from each other. We conclude that triple pulse TMS has a powerful facilitatory effect on MEP production under resting conditions and is highly dependent on the interpulse intervals. These results are consistent with the hypothesis that multiple TMS pulses can be used to target the properties of the cortical circuitry involved in the generation of I-waves.