This paper describes a methodology to develop a Computable General Equilibrium model with a sub-national detail starting from a global database and model presenting the country-level as the highest resolution. This procedure is demonstratively applied to Italy, but can be transferred to any country/macro-region, provided regional data availability. Increasing the spatial resolution of a CGE model can be particularly useful to capture local specificities not only in response to given policy shocks, but also to environmental impacts, as, for instance, those originated by climate change, which are highly differentiated spatially. Conceptual and practical issues are treated: we use an innovative method to estimate bilateral trade flows across sub-national areas and analyse the implications of different assumptions on both factor and good intra-country mobility. We carry out a simple experiment to test the robustness of our regionalized structure.