Sanctions are used ubiquitously to enforce obedience to social norms. However, recent field studies and laboratory experiments have demonstrated that cooperation is sometimes reduced when incentives meant to promote prosocial decisions are added to the environment. Although various explanations for this effect have been suggested, the neural foundations of the effect have not been fully explored. Using a modified trust game, we found that trustees reciprocate relatively less when facing sanction threats, and that the presence of sanctions significantly reduces trustee's brain activities involved in social reward valuation [in the ventromedial prefrontal cortex (VMPFC), lateral orbitofrontal cortex, and amygdala] while it simultaneously increases brain activities in the parietal cortex, which has been implicated in rational decision making. Moreover, we found that neural activity in a trustee's VMPFC area predicts her future level of cooperation under both sanction and no-sanction conditions, and that this predictive activity can be dynamically modulated by the presence of a sanction threat.cooperation ͉ neuroimaging ͉ perception shift ͉ punishment ͉ social norms S anctions are ubiquitous in modern human societies (1). The purpose of sanctions is to enforce norm obedience beyond the level that humans might achieve in the absence of punishment (2-4). However several recent field studies and laboratory experiments have established that adding monetary sanctions to an environment can reduce cooperation (5-7). Substantial speculation has arisen surrounding the source of this counterintuitive effect, including the possibility that the presence of sanctions might change individuals' perceptions of the environment, thus crowding out internal motivations for cooperation (5-8). The imposition of sanctions also might be perceived as a signal of distrust (9-11) and might create a hostile atmosphere (12, 13), leading to decreased cooperation.Previous behavioral experiments have sought to distinguish these competing explanations. For example, a recent study (5) reported data from an experiment aimed at determining the relative importance of intentions and incentives in producing noncooperative behavior. Participants played a one-shot investment experiment in pairs. Investors sent a certain amount to trustees, requested a return on that investment, and, in some treatments, could threaten sanctions to enforce their requests. Decisions by trustees facing threats imposed (or not) by investors were compared with decisions by trustees facing threats imposed (or not) by nature. The main finding was that when not threatened, trustees typically decided to return a positive amount less than the investor requested, but when threatened, that decision was less common. This result is the same whether the sanction is imposed by a human investor or by nature, suggesting that the detrimental effect of sanctions on cooperation might not hinge specifically on trustees' perceptions of investor intentions.