The project proposes to investigate the neural processes along with the contingencies that underlie the relation between eye cues and cooperative behavior in social dilemmas. We differentiate between real and stylized, (iconic) eye cues, and we propose that only the former provide social information and are capable of eliciting trust by affecting the brain's social cognition system (temporo-parietal junction, medial frontal cortex, and amygdala), while the latter, inducing a feeling of being watched, influence a person's reputation concerns by affecting neural activity in the lateral PFC. In addition, we expect that the effect of real versus stylized eyes (and underlying patterns of brain activation) should vary according to a person's social value orientation (SVO). We test these hypotheses in three different contexts, a simultaneously and sequentially played prisoner's dilemma game and an assurance game because these three games differ in their motives of greed and fear that typically drive people towards non-cooperation. We test if real versus stylized eye cues curb fear and greed respectively (thereby increasing cooperation), depending on a person's SVO.