This paper examines the relationship between daily travel distance and spatial characteristics in Flanders (and partly also in Brussels), in the north of Belgium. Important regional variations in commuting trip lengths are noticed, which are related to the spatial-economic structure including aspects of population density and spatial proximity. Commuter data obtained from the General Socio-Economic Survey 2001 are area covering and offer a lot of information. It is obvious that residents in the economic core areas produce less commuter mobility than people living in remote areas that have still access to the Brussels-Antwerp region. Proximity between home and work locations is paramount, when proximity is defined at a large geographical scale. Next, the spatial distribution of commuting distances, based on residential location, is compared to overall daily travel patterns including non-work travel. Since the second kind of data is only available in the form of a rather small sample, a multivariate regression equation based on spatial characteristics has been developed in order to extrapolate sample data throughout the Flanders region. When assessing overall daily travel patterns, including non-work travel, variables based on the spatial distribution of jobs do not show significant effects on the travel distance. However, spatial proximity is again paramount, although proximity should now be defined at a much smaller geographical scale. When considering all daily travel, the distance between the residence and an even small urban centre is much more decisive than the distance to the economic core areas (which is mainly consisting of the Brussels-Antwerp region). This finding qualifies the limited importance of the commute: today, it is mainly non-professional travel that is growing. Furthermore, the results suggest that residential density and land use mix in urban areas is the best guarantee for curbing excessive forms of overall (but especially: non-commuter) mobility.