A core claim of sociological theory is that modern institutions fall short of their meritocratic ideals, whereby rewards should be allocated based on achievement-related criteria. Instead, high-status actors often experience a “status advantage”: they are rewarded disproportionately to the quality of their performance. We develop and test a theory of status advantage in meritocratic settings. The most influential model in past research derives status advantage from decision-makers’ tendency to infer quality from status when quality is uncertain. The theory developed here integrates and extends this and other theories to explain the emergence of status advantage in the many meritocratic contexts where the decision-maker’s personal, first-order sense of quality is less important to the decision. We argue that in such contexts, decision-makers must often coordinate with others to make the “best” decision, and thus they focus on the “third-order inference” problem of discerning who or what “most people” think is higher quality, as encoded in status beliefs. Three experiments demonstrate that under such conditions, status advantages can emerge even though (1) status information does not resolve uncertainty about quality; (2) the status belief is illegitimate; and (3) no party to the decision personally prefers the higher-status option. The theory implies that status hierarchies are resilient in the face of significant dissent but may be subject to public challenge.