Thesis: Humanitarian and food aid have been widely criticized because of a lack of efficacy. This thesis investigates the behavior of each different actors involved in the process that could explained the relative efficacy of aid. The first part of the work investigates changes in decisions of production, sales and purchases induced by food aid. It looks at the extensive and intensive margin. The second chapter focuses on donors and implementing agencies. It documents humanitarian aid allocation and fragmentation at both levels. It investigates the potential consequences of delegation and fragmentation on humanitarian aid efficacy. Based on three case studies, the chapter shows that fragmentation is not necessary negative. The last chapter looks at strategic interactions between food aid donors using a reform implementing by the EU as a natural experiment.