This study proposes an alternative approach to estimate excess travel using activity diaries, extending it to non-work travel and capturing interhousehold variation. Central to this approach is a residential relocation exercise that helps to decompose the amount of travel by a household into two types of travel: required and excess travel. An empirical demonstration of the approach in North Carolina shows that on average 30 per cent of households' travel in the study area is in excess. Furthermore, lower levels of required and excess travel are found among households living in neighbourhoods with more compact development patterns, suggesting that households in such neighbourhoods are able to concentrate daily destinations in relatively smaller geographical areas and achieve better spatial co-ordination between residential and activity locations. Results also indicate that households make trade-offs between travel efficiency and neighbourhood amenities such as security and school quality, suggesting the importance of social dimensions in addressing transport problems.