Before it reaches the targeted beneficiary, a successful development assistance activity has to pass through the hands of many actors, including a slew of administrative and specialist individuals and committees within USAID, the implementing agent (which could be a government, private sector or non-profit agency), the recipient government political and administrative levels, and potential recipient country facilitators, to name but a few. Each of these actors has her own objectives and therefore idiosyncratic incentives, perhaps only incidental to those of USAID. On the other hand, many of these agents possess skills and information that could contribute to the success of the activity. How does one encourage these actors to enthusiastically provide their contributions without at the same time allow them to engage in potentially opportunistic behavior, which might derail the overall activity? Likewise, how can strategy, program, and activity designers at USAID not just avoid the pitfalls of these incentive conflicts and information asymmetries but explicitly harness them for the good of the intervention? We begin by describing the nature of these challenges. We then present a conceptual framework called Harnessing the Power of Incentives (or HPI, pronounced hippy). Based on the new institutional economics, it includes techniques that USAID may use both to reduce the risk of having their interventions hijacked or blunted by uncooperative collaborators as well as to turn these embedded incentives to their own advantage.The framework comprises several levels of analysis, each addressing different aspects of the aid program cycle. The macro level looks across activities and countries and identifies donor, recipient country, implementer, and activity characteristics that enhance the probability of a successful intervention or increase the likelihood of failure. The meso level, corresponding to the political economy dimension, identifies the interests of actor organizations and the games they play. With it, USAID can better manage its interactions with other organizations. The micro level elucidates the games and incentives facing the individual decision-makers within each organization and how these internal incentive structures may ultimately lead to stable but inefficient - and even perverse - aid outcomes. These three levels also provide guidance on improving incentives and collective action within and between USAID and its field missions and on the design of governance struc tures to improve implementer and USAID staff performance. Finally, we suggest the rudiments of an evaluation process comprising indicators and an experimental design, which would support aid effectiveness incentives at each of the three levels. We close by laying out a series of practical steps to implement an HPI application, with special attention to the different internal objectives and requirements of USAID administrative processes and their time and human resource constraints.