Monitoring conspecifics is a crucial process in social learning and a building block of social cognition. Selective attention to social stimuli results from interactions of subject and stimulus characteristics with dominance rank often emerging as an important predictor. We extend previous research by providing as stimuli naturally occurring affiliative interactions between group members instead of pictorial or auditory representations of conflicts, and by extending to the affiliative relationship, i.e. social bond, between subject and stimulus instead of just their dominance relations. Our observational data on adult female rhesus macaques support the prediction that subjects pay more attention to affiliative interactions of others than to solitary controls. Exceedingly more attention was paid to conflicts unfolding in the group which can have more prompt and direct consequences than others' friendly interactions. The valence of the stimulus (affiliative vs. agonistic) affected biases towards individuals dominant over the subject, but not the ubiquitous bias towards close affiliates of the subject. Keeping track of the whereabouts and interactions of key social partners has been proposed as a prerequisite for behavioral coordination among bonded partners. In groups of socially very active monkeys, social attention is gated by both social dominance and social bonding.Vertebrates living in social groups evolved a suite of socio-cognitive skills to maneuver their social environment 1 . These skills include individual recognition, tracking one's own past interactions, monitoring others' interactions, assessing others' relationships, and attributing mental states to others. Such abilities are thought to place increasingly large demands on the cognitive apparatus 1 . Processes involved in the use of social knowledge are a) the acquisition of social information, i.e. information on other individuals and their interactions, b) the application of such information, and c) its exploitation 2 . Each of these processes is subject to phenotypic constraints leading to inter-individual variation in the use of social knowledge 2 . There exists now ample evidence both for the ability to individually recognize conspecifics 3 and for gregarious animals to take into account past events when making social decisions 4 . Therefore, the focus of social cognition research shifted to understand variation in social attention, in social information acquisition, and in how and when individuals monitor others.Social information can function as a reward and incentive for nonhuman primates in experimental tasks suggesting that the acquisition of social information is important 5-7 . Not all information is valued to the same degree however, leading to selective attention varying with traits of the subject, of the stimulus individual or event, or their combination. Social information is thought to be valued particularly high, if its acquisition leads to direct benefits for the attentive subject. Selective attention to less immediate, indirect or ...