This study explores the potential of conjoint choice experiments for planning decisions on urban sites. We elicit people's preferences for regeneration projects that change the aesthetic and use character of specified urban sites. We use a split-sample design with two sets of regeneration projects. The first entails hypothetical transformations of an actual square with an important cultural and historical dimension. The other consists of hypothetical transformations of an abstract square which we try to make as close as possible to the former in all respects, except for its cultural and historical dimension. Each regeneration project is defined by aesthetic and use attributes. Our results suggest that individual choices are explained by the attributes, andthat the marginal utilities are significantly different across projects forthe actual and the abstract square.