Paired associative stimulation (PAS), combining transcranial magnetic stimulation (TMS) with electrical peripheral nerve stimulation (PNS) in pairs with an optimal interstimulus interval (ISI) in between, has been shown to influence the excitability of the motor cortex (MC) in humans. However, the underlying mechanisms remain unclear. This study was designed to explore an optimal protocol of PAS, which can modulate the excitability of MC in rats, and to investigate the underlying mechanisms. The resting motor thresholds (RMTs) of TMS-elicited motor evoked potentials (MEPs) recorded from the gastrocnemius muscle and the latency of P1 component of somatosensory evoked potentials (SEPs) induced by electrical tibial nerve stimulation were determined in male Sprague-Dawley rats (n=10). Sixty rats were then randomly divided into 3 groups: a PAS group (further divided into 10 subgroups at various ISIs calculated by using the latency of P1, n=5, respectively), a TMS (only) group (n=5) and a PNS (only) group (n=5). Ninety repetitions of PAS, TMS and PNS were administered to the rats in the 3 groups, respectively, at the frequency of 0.05 Hz and the intensity of TMS at 120% RMT and that of PNS at 6 mA. RMTs and motor evoked potentials' amplitude (MEPamp) were recorded before and immediately after the interventions. It was found that the MEPamp significantly decreased after PAS at ISI of 5 ms (P<0.05), while the MEPamp significantly increased after PAS at ISI of 15 ms, as compared with those before the intervention (P<0.05). However, the RMT did not change significantly after PAS at ISI of 5 ms or 15 ms (P>0.05). PAS at other ISIs as well as the sole use of TMS and PNS induced no remarkable changes in MEPamp and RMT. In conclusion, PAS can influence motor cortex excitability in rats. Neither TMS alone nor PNS alone shows significant effect.