Classic evolutionary and social exchange perspectives suggest that some people have more mate value than others because they possess desirable traits (e.g., attractiveness, status) that are intrinsic to the individual. This article broadens mate value in 2 ways to incorporate relational perspectives. First, close relationships research suggests an alternative measure of mate value: whether someone can provide a high quality relationship. Second, person perception research suggests that both trait-based and relationship quality measures of mate value should contain a mixture of target variance (i.e., consensus about targets, the classic conceptualization) and relationship variance (i.e., unique ratings of targets). In Study 1, participants described their personal conceptions of mate value and revealed themes consistent with classic and relational approaches. Study 2 used a social relations model blocked design to assess target and relationship variances in participants' romantic evaluations of opposite-sex classmates at the beginning and end of the semester. In Study 3, a one-with-many design documented target and relationship variances among long-term opposite-sex acquaintances. Results generally revealed more relationship variance than target variance; participants' romantic evaluations were more likely to be unique to a particular person rather than consensual. Furthermore, the relative dominance of relationship to target variance was stronger for relational measures of mate value (i.e., relationship quality projections) than classic trait-based measures (i.e., attractiveness, resources). Finally, consensus decreased as participants got to know one another better, and long-term acquaintances in Study 3 revealed enormous amounts of relationship variance. Implications for the evolutionary, close relationships, and person-perception literatures are discussed.