While the presence of direct anatomical connections cannot fully provide functional information, functional connectivity may indicate indirect connections. To noninvasively visualize the direct connections among distant regions, here we combined techniques of transcranial magnetic stimulation (TMS) and near-infrared spectroscopy (NIRS) for cortical stimulation and activation measurement, respectively. Based on our previous dynamic causal modeling study (Ohta et al., 2013), we focused on the direct connection from the left inferior frontal gyrus (L. IFG) to the left supramarginal gyrus (L. SMG). To examine selectivity of this connection, we also stimulated the left lateral premotor cortex (L. LPMC) and the right IFG (R. IFG), both of which may affect activation in the ipsilateral SMG. By applying twin-pulsed TMS (interstimulus interval, 2 ms), we stimulated each target region randomly at 18-24 sec intervals, repeating 20 times in a single run (Real condition). As a reference condition, the TMS coil was positioned 3 cm above the scalp (Sham condition). We tested two runs under each of Real and Sham conditions, the order of which was counterbalanced among the nine participants. For each event of NIRS measurements, we corrected the baseline by fitting a cubic line, and converted the corrected data to oxy- and deoxy-hemoglobin concentrations. Finally, we calculated the t-maps of the Real - Sham contrast for the same stimulated region, using NIRS-SPM software (http://bispl.weebly.com/nirs-spm.html#/). We found robust activation in the L. SMG when stimulating at the L. IFG (FWE-corrected p < 0.05), and the activated region due to the L. LPMC was more dorsally restricted in the L. SMG. In contrast, we observed no significant activation in the R. SMG when stimulating at the R. IFG. This clear lateralization effect indicates the presence of callosal inhibition from left to right regions, either in the IFG or SMG.