Building on the OECD’s Better Life Initiative and new work using geospatial analysis, this paper investigates how reported life satisfaction relates to some of the urban structure indicators. To this end, it merges OECD household survey data with urban structure data from OECD’s Metropolitan Database, which includes a number of city-level indicators such as population and road density, as well as localised measures of land-use. The merged data permit analysis for five countries: France, Japan, the Netherlands, Spain and Sweden. The findings from this analysis provide some evidence of a trade-off between home size and distance to the city centre, although the statistical power of this effect is relatively weak. Interestingly, regression analysis suggests that overall city-level compactness has a clear negative relationship with life satisfaction, regardless of whether individuals live in the urban core or in peri-urban areas. Land-use fragmentation is also found to have a negative relationship with individuals’ life satisfaction. These general patterns are for the most part robust to various statistical tests. They also hold when econometric analysis is conducted at the country level. Residents of cities with greater levels of centralisation – i.e. a greater share of the population living in the city centre – exhibit measurably lower levels of life satisfaction. A naïve interpretation of this result would suggest that anti-sprawl policies do not in fact improve overall welfare. This study does not support this conclusion. It does, however, give cause for consideration before accepting ‘win-win’ arguments for ‘smart growth,’ often brought forward to support increasingly concentrated, high-density development. The evidence presented here suggests that such policies are not without their welfare trade-offs, and that there will be winners and losers from their implementation. While high-density policies can clearly make a positive contribution to reducing local and global environmental externalities, many of these benefits are deferred and may largely accrue to future generations. A key general lesson from this study is that compensation of the losers may improve the equity effects of these policies, as well as prove more expeditious from a political economy perspective. One of the simplest approaches to compensation would be to balance pecuniary incentives for smart growth, such as higher development taxes or fees, with compensatory policies, such as subsidies or tax or fee offsets in other domains. The main policy conclusion from this study is that smart growth policies should include distributional analysis and recommendations for addressing concerns about inequalities flowing from the scoping and implementation of policies.