Accurate simulation of the city energy balance requires studying the thermal behaviour of buildings and therefore requires knowing many details: at least the buildings geometries, their envelope materials and surroundings ground covering. For instance, information on shape and location of a building are useful to assess the performance of solar panels; information on the materiality of the facades, make it possible to estimate solar gains through the windows and therefore to perform building energy balance simulations (Masson, 2000); etc. Consequently, several recent research works have been aiming to integer building scale data into urban scale simulations (microclimate, energy consumption of buildings at the city scale, energy production, etc.). Among those researches, we can quote the WUDAPT project (Ching, 2012) aiming to define a worldwide building database based on the LCZ classification of urban forms (Stewart & Oke, 2009). 2. Problematics The main issue in integrating this type of information to urban scale simulations is the lack of precision of the available data for buildings (Ching et al., 2009). If a limited number of buildings can be very precisely described (through existing Building Information Model (BIM) for instance (Ferries et al., 2014), but also through historical studies or architectural inventories), the data at the city scale remains broadly heterogeneous. In this paper, we will present how we used those localized descriptions of buildings to enrich existing urban database existing at national scale in the context of the MApUCE project (a French research program that aims to integrate quantitative data from urban microclimate, climate and energy in urban policies). 3. Method Our working method, called GENIUS (GENerator of Interactive Urban blockS), was to perform a literature review combined with interviews of urban planners to characterize a typology of urban forms in the whole French territory, and to associate it with a wide database. The key theme of our work concerns the integration of building scale data into urban microclimate and energy consumption modelling. It leads us to formulate the question "how can we characterize the building scale in oder to update, and to make full use of urban simulation tools ?" To answer this question, we have broken down our work into three consecutive sections:-The first section interview urban planners about the differents urban typologies in France. Our aim is to take in account the vision of designers operating processes, in order to identify a system of ranking of urban typology .-The second section covers existing buildings database for identify, characterize intangible datas, such as buildings use, date of construction and buildings location. The aim of this analysis is to create the conditions for identify representative buildings in France.-The third section defines the building scale data. It is supported by a bibliographical study of building materials and systems, architecture and building cultures, conservation, etc. and is aimed at reaching an understanding of how information on building scale can be used for urban simulation tools. To conclude, we define different paths to be developed with the aim of improving the link between scale building data and urban microclimate and energy consumption modelling. .