Most Western national governments aim to influence individual travel patterns - at least to some degree - through the spatial planning of residential areas. Nevertheless, the extent to which the characteristics of the built environment influence travel behaviour remains the subject of debate among travel behaviour researchers. This work addresses the role of residential-self-selection, an important issue within this debate. Households may not only adjust their travel behaviour to the built environment where they live, but they may also choose a residential location that corresponds to their travel-related attitudes. The empirical analysis in this work is based on data collected through an internet survey and a GPS-based survey, both of which were conducted among homeowners in three centrally located municipalities in the Netherlands. The study showed that residential self-selection has some limited effect on the relationship between distances to activity locations and travel mode use and daily kilometres travelled. The results also indicate that the inclusion of attitudes can help to detecting residential self-selection, provided that studies comply with several preconditions, such as the inclusion of the 'reversed' influence of behaviour on attitudes.