Learning to trust the right partner is pivotal to survival. But what information matters to decide whom to trust? In this chapter, we review evidence suggesting that different character traits play a role in the formation of trustworthiness impressions and beliefs that guide trust decisions. Learning of these traits depends on available information about the other person, previous knowledge, and contextual circumstances. Interestingly, when these factors favor the learning of particular traits, the resulting beliefs are harder to revise and lead to behavioral patterns that suggest a learning impairment. Computational models indicate an asymmetry in feedback valuation that is not due to the type of feedback (e.g., positive or negative) but rather to previous knowledge and contextual factors (e.g., the reputation of the other person). Neuroimaging studies highlight the role of mentalizing brain regions in building adequate mental models of others. Specifically, the orbitofrontal cortex and temporoparietal junction are central to the formation and updating of trustworthiness beliefs. Further, the dorsal posterior cingulate cortex and lateral frontoparietal regions likely underpin information integration processes for behavior change in face of untrustworthiness. We finally call for collaborative efforts in future scientific enterprises to develop a still lacking neurocomputational theory of social learning.