People's beliefs about how other people perceive their personality tend to be fairly accurate, but how does accuracy arise? The current research answers this question by testing three potential sources of meta-accuracy: the person forming the metaperception (i.e., the metaperceiver), the person forming a judgment about the metaperceiver (i.e., the perceiver), and the unique relationship between the two individuals (i.e., the dyad). In three studies, participants interacted with new acquaintances one-on-one in a platonic (N = 547) or dating setting (N = 378), or in a platonic group setting over time (Time 1, N = 242; Time 2, N = 191). Metaperceivers tended to have the most robust influence on meta-accuracy, but perceivers and especially dyads influenced accuracy as well. This suggests there are "good" metaperceivers, perceivers, and dyads of meta-accuracy and that a more complete understanding of meta-accuracy must consider both members of an interaction. As a first step in understanding how both individuals influence accuracy, we tested the role of self-perception, specifically if some metaperceivers, perceivers, or dyads fostered accuracy because metaperceivers happened to be seen as they saw themselves. Perceivers largely fostered accuracy by seeing metaperceivers as they saw themselves but metaperceivers and dyads mostly fostered accuracy by other means. Potential contextual effects are discussed. (PsycInfo Database Record (c) 2021 APA, all rights reserved).