Over the past fifteen years, the common agricultural policy (CAP) is profoundly changed under the pressure of European society and economic trends. The reforms of 2003-2004 marked a new phase in this path. This process continued with the reforms of the sugar (2006) and fruit and vegetables(2007) and is still ongoing with the recent proposals to reform the wine sector. The dynamics of consolidation, evaluation and redirection of the legislative measures adopted are very important and topical in view of the definition of the common agricultural policy. These issues on the political side emphasize the need for policy makers to have evaluative instruments witch can analyze both the effects on farmers behaviour and the effects generated in the territory. Using instrument that measures and localises the likely effects of agricultural policies, as a matter of fact, allows a correct and comprehensive analysis of actions taken, with the aim of bringing any corrections and adjustments to particular situation. The present work aims to implement and to test in a case study an evaluation instrument based on the integration of two different methodologies: the mathematical programming and the geographic information system. With this research we want to propose a possible method to study at the same time the behaviour of farmers and the localization of their choices under different political conditions. Farmers behaviour has been studied through the implementation of a positive mathematical programming (PMP) model. The construction of a georeferred database made possible to consider the spatial characteristics as variables directly involved in the process of mathematics optimization. Therefore, the geographic information system was used not only to locate the effects of policies, but also to constrain the mathematical model. The proposed model was applied in a rural area in the province of Florence, the Mugello. Through the comparison of different policy scenarios were identified and located the main effects induced by the mid term reform on soil, on livestock and on farms economic performance. Results obtained show how Cap reform induced a reallocation of farms resources and that it did not lead to the feared abandonment of primary activity. In addition, the elaboration of results provide important informations to decision public in order to develop monitoring actions in those areas which show potential problems. This model, for its flexibility and ductility, may constitute the basis for simulate and analyze the potential effects induced by the proposals, for example, in the health check of the CAP.