Group walking is a collective social interaction task as pedestrians are required to determine their own pace of walking on the basis of surrounding others’ states. The steady beat sound is known to be a controllable factor that contributes to relative success/failure of coordinated group walking since the beat improves pedestrian flow in congested situation. According to some reports, inter-personal interaction synchronizes inter-personal brain activity in the prefrontal region, which supports social cognitive processes required for successful inter-individual coordination, such as predicting each other’s state; success/failure of a coordinated task is associated with increase/decrease in inter-subject neural synchrony (INS). Combining these previous findings, we hypothesized that INS during group walking in congested situations would also differ depending on the existence of the steady beat, corresponding to the modulated coordination-related cognitive processes. Subjects’ frontopolar activities were measured using ultra-small near infrared spectroscopy, which can simultaneously measure the brain activities of multiple subjects without constraints on their motions. To exclude the possibility that increased INS may be spuriously induced by the shared stimuli (i.e., steady beat) or by the resultant behavioral synchronization, as control we used stepping on a same spot, which is similar in movement to walking but does not require the subjects to consider others’ states, either with or without the steady beat. In a two by two repeated measures factorial experimental design, the subjects were instructed to walk or keep stepping on a same spot with or without a steady beat sound of 70 beats per minute. As previously reported, the walking flow during group walking with the beat significantly increased compared with that without the beat. Synchronization of stepping between the subjects was also significantly increased by the steady beat sound. For INS, we observed a significant interaction effect between walking/stepping and sound/no-sound, supporting our hypothesis. INS while walking with the beat was higher than that without the beat, whereas the beat induced no significant differences in INS during stepping. Furthermore, the effect of the beat on INS while walking was spatially extended beyond the adjacent pedestrians, reflecting the diffuse nature of the collective coordination in group walking. The increase of INS for walking suggested that the steady beat sound led to more harmonized inter-personal cognitive processes, which resulted in the more coordinated group motion.