It is a widely held belief that civil development programs play a central role in any counterinsurgency campaign. It is assumed that civil development assistance is key to ‘winning the hearts and minds’ of the civilian population, which, in turn, is crucial for victory. However, there is currently little evidence to support this belief. This article by Dr Adam Lockyer begins by analysing the different methods that have been used in Afghanistan in order to evaluate the effectiveness of civil development programs since 2001. It finds that these methods have severe limitations. Indeed, based solely upon current methods of evaluation, we have no reliable evidence whether civil development programs are actively improving security, having no impact or making matters worse. As such, this article makes the case for a field experiment approach to be adopted in Afghanistan. It argues that field experiments are the most powerful methodology currently available to social scientists for making causal inferences and, by making minor changes to the way in which civil development is distributed, we can vastly improve our understanding of the relationship between aid and security.