In most birds and mammals, young are raised in family groups. The phenotypes of nestmates and parents are thus reliable cues for recognition of conspecifics and kin. However, in some species, young develop alone, or in broods of mixed relatedness (e.g. because of multiple paternity or maternity), or among heterospecifics or unrelated conspecifics (brood parasites). Under these circumstances, the best referent (model) for discriminating close from distant kin and heterospecifics from conspecifics might be one's own self. This recognition process is known as self-referent phenotype matching. Here we review recent experimental evidence of self-referencing and suggest that behavioral neuroscience can provide new tools and insights into how it works (its proximate mechanistic and ontogenetic bases) and why it exists (its adaptive significance).