For over 100 years, a central project of administrative law has been to legitimize administrative agencies within the constitutional order. Scholars and judges assert that various structural features and administrative procedures—e.g., expertise-driven decision-making, channels for public participation, and presidential control—enhance agencies’ “legitimacy.” Although these authors virtually never define the term, most definitions of legitimacy—from Weber and Habermas to contemporary legal scholars—involve some measure of acceptance by people outside of narrow legal elites. Accordingly, whether a given design feature fosters legitimacy is in part an empirical question. I propose a set of experiments to gauge respondents’ reactions to various features that are claimed to legitimize agencies. Respondents will read a short vignette presenting a policy decision facing an agency, describing the process by which the agency made a decision, and then stating the eventual decision. The structural features and procedures included in the description will vary by respondent. Respondents will then be asked their views on the agency and its decision. The goal is to assess competing claims regarding which elements of administrative law, if any, buttress the administrative state’s legitimacy