This chapter reviews the different methods that are used in impact evaluation and illustrates applications of these methods to various development interventions. Impact evaluation of public programs is often required by law. This is the case in Mexico, where all programs that use federal resources are expected to report impacts on a bi-annual basis to the executive and the legislative bodies. Impact evaluations have three main objectives: accountability, results-based management, and generic lessons for the profession. A Randomized Control Trial (RCT) requires that the organization selects more candidate villages or schools than the number in which it would be willing to intervene and then randomly selects from among these units the subsample that will receive the intervention. It is possible to find natural experiments that mimic an RCT. While an RCT may be the cleanest way of organizing an impact evaluation, several difficulties may arise.