The knowledge base of empirical psychology is more expansive than ever before. So too is the impulse to integrate this factual knowledge into political theory. But how should this psychological turn be undertaken? What would a political psychology for political theorists look like? How could psychology credibly tackle the questions that political theorists characteristically ask, especially regarding the nature and consequences of prescriptive political judgment? In this thesis, I explore this issue through the framework of recent debates between political moralists—specifically, John Rawls, G. A. Cohen, and Peter Singer—and political realists—largely Bernard Williams. Deploying the insights of political realists, I argue that moralists cannot quarantine the relevance of psychological facts through the ideal of a 'pure' normative judgment. To explore what this empirical engagement might look like, I contrast these moralist ideals of judgment with Jonathan Haidt’s social intuitionism, which proposes a more affectively laden and pluralistic model of judgment. I then redeploy the insights of political realism to critique social intuitionism, to uncover its weaknesses from the perspective of existing political theory. Finally, to stabilize this critique, I lay out the framework for a reflexive political psychology, which acknowledges the co-constitutive relationship between the discipline of psychology and its subject matter: human psychology. This reflexive political psychology offers an agenda by which we can investigate the political usefulness of psychological and political theories.