In the past two decades human geographers have intensely theorized scale, and extended claims that it is a foundational element of geographic theory. Yet attendant with this move has been a growing concern that scale has become an unwieldy concept laden with multiple, contradictory and problematic meanings. 1 share that concern, and argue that a similar debate about the usefulness of 'identity' as a conceptual category in social science offers instructive insights. Paralleling recent critiques of identity categories such as nation and race, I view the conceptual confusion surrounding scale and scale politics as, in part, the consequence of failing to make a clear distinction between scale as a category of practice and category of analysis. In adopting scale as a category of analysis geographers tend to reify it as a fundamental ontological entity, thereby treating a social category employed in the practice of sociospatial politics as a central theoretical tool. I argue that this analytical manoeuvre is neither helpful nor necessary, and outline its consequences in analyses of the politics of scale. Finally,1 sketch the altered contours of a research programme for the politics of scale if we take this injunction seriously both in terms of how we theorize scale as a category of practice and what becomes the focus of scale politics research.