Forming judgments of other people is one of the most basic and consequential elements of social life, impacting virtually every interpersonal and intergroup interaction. The impressions we make of others comprise the integration of many, often varied, attributes about the person, including cues to their social group membership, traits of their personality we believe them to possess, and characteristics of their behavior we witness. The theories of person perception that explain this integration of attributes have consistently relied upon a core set of claims: a) the processing of certain attributes (e.g., individuating behaviors) are affected by cognitive capacity more than the processing of other attributes (e.g., social categories); b) accuracy and impression management motivations should reduce the use of irrelevant and/or biasing attributes (e.g., social categories cuing stereotypes); and c) attributes are processed in competition, such that greater use of one attribute results in less use of another attribute. However, current measures of person perception have not afforded appropriate tests of these claims. Using a formal modeling approach that avoids previous limitations to measuring person perception, we conduct rigorous tests of all three theoretical pillars to person perception research. Our results contend with those pillars and provide both novel insights to advancing both the measurement and theoretical understanding of social judgments.