A large share of a neighborhood project’s environmental impacts is due to mobility. It either takes place inside the neighborhood, such as transit traffic or internal mobility, or is induced by it and exchanged with the rest of the urban area. A way to improve mobility impacts evaluation in the assessment of neighborhood alternative designs, is to refine traffic simulation models making them more sensitive to spatial design while keeping their sensitivity to local traffic conditions and associated energy consumption and pollutants’ emissions. This paper introduces a methodology relying on the classic four-step scheme for mobility demand modelling together with specific spatial refinement. The neighborhood is divided into fine sub-areas, with specific consequences for each step: first, trips are generated on the basis of sub-area land-use and activity data; second, the trips are distributed between all Traffic Analysis Zones (TAZs), enabling to identify internal short-range trips; third, the mode choice model takes into account the particular access conditions between sub-areas and transit stations or roadway nodes; fourth, traffic assignment involves finer TAZs and finer path description. Furthermore, a 5th step is added to deal with environmental evaluation, especially the allocation of mobility impacts to the project’s sub-areas. These steps are presented and illustrated on the ‘Cité Descartes’ district case study, in Eastern Paris. Dividing its 1 km² area into about 100 sub-areas enabled us to depict the projects’ program and spatial layout very finely, especially so in relation to the transit stops and stations location. Some limitations and needs for further research are also outlined.