This paper addresses the problem of measuring neighbourhood characteristics and change when working with individual level datasets to understand the effects of residential mobility. Currently available measures in Britain are in various respects unsuitable for this purpose. The paper explores a new indicator of small area poverty: the Unadjusted Means-tested Benefits Rate (UMBR), which divides claimants of means-tested benefits in a small area by the number of households. We describe changes in area poverty between 2001 and 2006, using UMBR. As often assumed, these are generally negligible, but small areas in Bdisadvantaged urban^and Bmulticultural city life^communities did change considerably in this period. We also link UMBR to the first three waves of the UK Millennium Cohort Study, a survey of families with children born at the beginning of the 2000s. We examine opinions about Appl. Spatial Analysis (2016) neighbourhood and find that parents living in areas of higher poverty did tend to express more negative views than those living elsewhere. Living in high poverty areas was also associated with moving home, and those families who retrospectively gave neighbourhood considerations as reasons for moving did move into areas with markedly lower poverty rates. Finally, we compare families' moving trajectories to trends in poverty within areas. We are able to show that a large proportion of families who moved to poorer neighbourhoods were at double disadvantage, as they often moved to areas with increasing poverty rates. We conclude that UMBR can be used to enhance understanding of changing neighbourhood contexts in cohort studies, at least for this period, although it still suffers from the same conceptual and technical difficulties as other available alternatives in terms of its ability to capture aspects of neighbourhood quality.