One of the most important assumptions of kin selection theory is that individuals behave differently towards kin than non-kin. In mammals, there is strong evidence that maternal kin are distinguished from non-kin via familiarity. However, little is known about whether or not mammals can also recognize paternal kin as many female mammals, including primates, mate with multiple males near the time of conception, potentially concealing paternal kinship. Genetic data in several mammalian species with a promiscuous mating system and male-biased dispersal reveal a high skew in male reproduction which leads to co-residing paternal half-siblings. In most primates, individuals also form stable bisexual groups creating opportunities for males to interact with their offspring. Here I consider close paternal kin co-resident in the same social group, such as father-offspring and paternal half-siblings (i.e. animals sharing the same father but who were born to different mothers) and review mammalian studies of paternal kin discrimination. Furthermore, I summarize the most likely mechanisms of paternal kin discrimination (familiarity and phenotype matching). When familiarity is the underlying mechanism, mothers and/or the sire could mediate familiarity among paternal half-siblings as well as between fathers and offspring assuming mothers and/or fathers can assess paternity. When animals use phenotype matching, they might use their fathers' template (when the father is present) or self (when the father is absent) to assess paternal kinship in others. Available evidence suggests that familiarity and phenotype matching might be used for paternal kin discrimination and that both mechanisms might apply to a wide range of social mammals characterized by a high skew in male reproduction and co-residence of paternal kin. Among primates, suggested evidence for phenotype matching can often have an alternative explanation, which emphasizes the crucial importance of controlling for familiarity as a potential confounding variable. However, the mechanism/s used to identify paternal kin might differ within a species (as a function of each individual's specific circumstances) as well as among species (depending upon the key sensory modalities of the species considered). Finally, I discuss the possible cues used in paternal kin discrimination and offer suggestions for future studies.