s / Brain Stimulation 8 (2015) 412e427 416 Understanding the actions of other people is a crucial skill to function in a social environment. When observing actions, one can understand implied goals and mental-states. The mirror neuron system (MNS) plays an important role in action understanding by internal simulation of the observed movements. This simulation process may be sufficient to decipher thoughts and intentions incorporated in an action. However, a more abstract mentalizing network may be necessary to efficiently process these inner states. In a recent fMRI study, we demonstrated that nodes of the theory-of-mind network (ToM) are activated when participants observe hesitations conveyed in hand movements. However, activation of a brain region does not the imply its necessity to perform the given task. Therefore, we designed a repetitive TMS study using the same experimental paradigm. On three experimental days, participants received 15 minutes of 1-Hz inhibitory rTMS stimulation, either on a MNS node (PF in the inferior parietal lobe), a ToM node (temporo-parietal junction (TPJ)), or a control region (vertex). After the stimulation, the participants performed three tasks: judging the size of a chosen object based on hand aperture, judging the level of hesitation and a control task. The first hypothesis was that a virtual lesion of PF would impair the performance in the hand aperture and hesitation task because they require the analysis of movement nuances. The second hypothesis was that a virtual lesion of the TPJ would influence the performance on the hesitation task only, because it requires inference about a mental state (confidence of choice). As predicted, we observed increased response latencies on the hand aperture task following the PF stimulation and increased response latencies on the hesitation task following both PF and TPJ stimulation. These results suggest that the TPJ is necessary to efficiently decipher mental states encoded in hand actions.