Abstract: Policies on spatial planning and on mobility often take it for granted that a location policy that favours situating human activities near junctions of public transport results in a frequent use of sustainable modes of transport. But in the daily practice of mobility planning we come across more and more critical statements. As the Flemish Ministry of Mobility and Public Works considers investing resources in an improved location policy, they asked for evidence on the possible results of location policy on sustainable commuting in a Flemish context, including Brussels. Quite exceptionally, we utilise the individual census data from nearly all 1, 2 million Flemish commuters. These detailed data allow a precise description of the commuter characteristics of different locations and allow the construction of a robust geographical pattern. We focus the analysis on four kinds of locations: railway stations, public transport junctions, urban areas and areas with a high economic density. An exploratory data analysis suggests that people working and/or living in these areas travel less distance, especially by car, and make more use of public transport and slow modes. A cluster analysis shows that there are spatial differences, resulting in clear geographical patterns, which suggest that policies should be spatially differentiated. As a result, we support the idea of using location policy to produce more sustainable commuting. From a methodological point of view, we conclude that census data are irreplaceable to meet the requirements of spatial representativity needed to construct detailed geographical patterns of commuting.