This paper explores associations between local socioeconomic context and perceptions of neighbourhood quality of life, using a multilevel modelling approach to analyse data from the 2002/03 Survey of English Housing. A range of individual, property and area variables is successively introduced, the aim being to determine which of these has the greatest influence on the probability of being dissatisfied with one's neighbourhood. Several variables relating to property type and characteristics were derived and a set of ecological independent variables was used that is different from those deployed in previous analyses. Using a multilevel approach, the analysis determines the spatial scale at which variance is most concentrated—namely, the primary sampling unit (PSU). The relationship between tenure and property type is broken down and it is shown that social renting per se is not necessarily associated with greater dissatisfaction, although there is an association between private renting and greater satisfaction, probably reflecting the age and social profile of private renters. In respect of ecological indicators, increased levels of deprivation were associated with a raised probability of expressing neighbourhood satisfaction, as was settlement size, which is consistent with earlier work on quality of life. Conclusions are drawn concerning the use of studies of this kind to monitor trends in neighbourhood satisfaction.