Public law is pervaded by anxiety about the role of courts. This is implicated perhaps most acutely in doctrines of non-justiciability, whereby courts decline to adjudicate issues on the basis that they raise questions—sometimes dubbed ‘political’—that are alleged to be unsuitable for judicial resolution. Amidst the many critiques of such doctrines, there is little attention paid to the prior question of what it is to hold a question non-justiciable in this sense. This talk will offer an account of the conceptual structure of non-justiciability in the hope of illuminating its normative structure. I distinguish two types of non-justiciability, argue that they have different normative implications, and make some suggestions about when (if at all) each should be deployed.