Based on H-reflex data, spinal mechanisms are proposed to be responsible for the first 50-80 ms of the transcranial magnetic stimulation (TMS)-induced silent period. As several methodological issues can compromise H-reflex validity as a measure of motoneuron excitability, this study used transmastoid stimulation to elicit cervicomedullary motor evoked potentials (CMEPs) during the silent period. Eleven subjects made 1-3 visits which involved 32 or 44 brief (~3 s) isometric elbow flexor contractions at 25 % of maximal torque. During each contraction, transmastoid stimulation was delivered in isolation to elicit an unconditioned CMEP and at interstimulus intervals (ISIs) ranging from 50 to 150 ms after TMS to elicit a conditioned CMEP. Stimulus intensities for TMS and transmastoid stimulation were set to elicit a silent period of ~200 ms and an unconditioned CMEP of 15, 50, or 85 % of the maximal compound muscle action potential (M ), respectively. At all ISIs and intensities of transmastoid stimulation, the conditioned CMEP was significantly smaller than the unconditioned CMEP (p< 0.001). However, suppression of the conditioned CMEP was significantly less at 85 % compared to 15 or 50 % M (p = 0.001). Contrary to published H-reflex data, the conditioned CMEP did not recover within 50-80 ms, remaining significantly suppressed at the longest ISI tested (150 ms). These data suggest the spinal portion of the TMS-evoked silent period is considerably longer than reported previously. Transmastoid stimulation, unlike peripheral nerve stimulation, does not impact proprioceptive inflow to motoneurons. Hence, relative to the H-reflex, the CMEP will be subjected to greater afferent-mediated disfacilitation and inhibition due to the TMS-induced muscle twitch.