Drawing on a database of more than one hundred anti-poverty interventions in 47 countries, this report provides a general review of experiences with methods used to target interventions in transition and developing countries. Written for policymakers and program managers in developing countries, in donor agencies, and in nongovernmental organizations who have responsibility for designing interventions that reach the poor, it conveys what targeting options are available, what results can be expected as well as information that will assist in choosing among them and in their implementation. Key messages are: 1) While targeting "works" - the median program transfers 25 percent more to the poor than would a universal allocation - targeting performance around the world is highly variable. 2) Means testing, geographic targeting, and self-selection based on a work requirement are the most robustly progressive methods. Proxy means testing, community-based selection of individuals and demographic targeting to children show good results on average, but with considerable variation. Demographic targeting to the elderly, community bidding, and self-selection based on consumption show limited potential for good targeting. 3) There is no single preferred method for all types of programs or all country contexts. Successful targeting depends critically on how a method is implemented.