This paper takes stock of Madagascar's first 10 years of decentralization. As it happened in many other developing countries, particularly in Africa, Madagascar's decentralization process has seen reversals, uncertainties and lack of clarity all along. This explains why Madagascar, despite the experience with decentralization, remains a highly centralized country with only about 3-4 percent of expenditures spent below the center and with very few prerogatives decentralized to the local level. Notwithstanding the structural impediments to decentralization in poor countries, many positive lessons can be drawn from the Madagascar case, which point to the potentials of the decentralization process. This study provides a detailed analysis of local government finances and develops a methodology for measuring local financing needs (local fiscal gap methodology). Based on this analysis, the study argues that a lot can be gained from simplifying administrative arrangements and fiscal relationships. Instead of a full-blown and ambitious decentralization strategy, this book suggests a number of reforms, which would go a long way by making the current structure work better. These reforms include: (1) a full transfer of the (limited) local competencies to commune, particularly local revenue collection; (2) increasing transfers to rural communes so that per capita allocations would be the same across communes-rural and urban; and (3) assigning revenues to one level of government only, except for some very specific types of taxes (such as on natural resources).