In this thesis two different approaches for the construction of environmental extensions compatible with the MRIO database EUREGIO are compared alongside with the latest liter- ature on consumption-based accounts. To date, few studies have focused on the provision of sub-national accounts at the NUTS2 level of detail and never production-based accounts have been compiled for the European NUTS2 level of detail. The first approach described as the bottom-up approach uses TNO’s MACC emission inventory for the year 2010 which compiles spatially explicit emission data at the source for 9 different pollutants across Europe with a 0.1àx 0.05à longitude-latitude resolution for over 250 source sectors. The second approach described as the top-down approach uses NAMEA data from EUROSTAT by disaggregating it to the regional level following a linear approach based on the regional economic output of a sector compared to the national economic output of the same sector. The two differ- ent approaches for the compilation of environmental extensions are compared both at the production-based accounts and the consumption-based accounts level mapping the existing relative difference across sectors at the national and regional level. The level of heterogeneity in the magnitude of the relative difference of the accounts increases proportionally with geo- graphic detail underlying the fact that a bottom-up construction is able to better grasp the regional scale. This is ever more prominent when looking at particular sectors such as the utilities sectors ss2 and ss5. Of the analysed NUTS2 European regions, the carbon footprint of 74% of the regions presents an upwards deviation of more than 4% compared to the bench- mark study with footprints ranging from 2 to 46 tCO2/cap for Umbria (IT) and Souther Finland (FI) respectively. The obtained results highlight the advantage of the bottom-up approach for more accurate distribution of the emissions at the regional scale from industries to the consumers resonating with similar conclusions from other authors. Despite this, several recommendations are made to improve the quality of the bottom-up approach with possible future applications.