A growing body of literature discusses the CO2 emissions of cities. Still, little is known about emission patterns across density gradients from remote rural places to highly urbanized areas, the drivers behind those emission patterns and the global emissions triggered by consumption in human settlements—referred to here as the carbon footprint. In this letter we use a hybrid method for estimating the carbon footprints of cities and other human settlements in the UK explicitly linking global supply chains to local consumption activities and associated lifestyles. This analysis comprises all areas in the UK, whether rural or urban. We compare our consumption-based results with extended territorial CO2 emission estimates and analyse the driving forces that determine the carbon footprint of human settlements in the UK. Our results show that 90% of the human settlements in the UK are net importers of CO2 emissions. Consumption-based CO2 emissions are much more homogeneous than extended territorial emissions. Both the highest and lowest carbon footprints can be found in urban areas, but the carbon footprint is consistently higher relative to extended territorial CO2 emissions in urban as opposed to rural settlement types. The impact of high or low density living remains limited; instead, carbon footprints can be comparatively high or low across density gradients depending on the location-specific socio-demographic, infrastructural and geographic characteristics of the area under consideration. We show that the carbon footprint of cities and other human settlements in the UK is mainly determined by socio-economic rather than geographic and infrastructural drivers at the spatial aggregation of our analysis. It increases with growing income, education and car ownership as well as decreasing household size. Income is not more important than most other socio-economic determinants of the carbon footprint. Possibly, the relationship between lifestyles and infrastructure only impacts carbon footprints significantly at higher spatial granularity.