The built environment affects our lifestyles and associated carbon emissions. Most studies investigating this link, however, have used census data or surveys that can hardly capture patterns of daily activities over time. In this paper, we present the results of an eight-week monitoring campaign during which over 300 residents of the Province of Utrecht (Netherlands) recorded their activities pertaining to the domains of travel, housing energy and food consumption using a mobile app. Data were analyzed through structural equation models (SEMs) accounting for residential self-selection. Once controlling for socio-demographics, an aggregate measure of compactness encompassing population density, intensity of development and accessibility of transit and grocery stores was found to be associated with emissions in all three domains via multiple mediating variables. More compact neighborhoods may curb emissions by reducing car ownership and therefore driving, but seem to have no impact on other travel habits. Yet, they tend to be associated with residential settings (i.e. older multi-dwelling units) showing higher energy emission factors due to less reliance on green sources. Additionally, lower rates of car ownership in more compact neighborhoods are associated with less carbon-intensive diets. While our findings endorse the compact city as an instrument to help cut individual carbon emissions, they also raise warnings about the ability of low-carbon technologies (e.g. electric vehicles, photovoltaic systems) to make this spatial planning model look obsolete in the fight against climate change. Mobile apps like ours may help disentangle this puzzle though greater automation is needed to get unbiased information.