The movement for localization in foreign aid presents philosophical and practical challenges for providers of foreign aid. In the past, some level of localization (to at least state actors) of both decision-making and spending could be achieved through general and sector budget support, a logistically simple (though politically difficult) approach. Current practices, with more aid disbursed in the form of discrete projects, make localization more difficult to achieve. For development agencies to make progress they must confront conceptual, practical and strategic challenges. This paper suggests practical definitions of different kinds of localization and what they seek to achieve, points out trade-offs between them and sets out how they challenge current conceptions of how and what foreign aid seeks to do, as well as suggest limits to the optimal extent and type of localization. It then considers practical difficulties in implementing these different conceptions of localization given existing donor practices and capacities, and the strategic challenges of realizing the desired outcomes—specifically principal-agent problems within the organization and with contractors. It concludes with key lessons and messages for donors and funding agencies.