The causal relationship between urban form, in particular density, and travel demand is subject to debate. Here, we investigate this relationship using a structured regression approach applied to a large-scale layered dataset of travel patterns and urban form. We find that residents of the densest urban areas use 80% less energy for transportation than residents of rural areas and explore the causal factors influencing this relationship. We find that the primary driver of the density–energy use relationship is the larger number of nearby destinations available to urban dwellers, followed by differences in road network properties and public transit infrastructure. The energy benefits of urban dwelling are weakened when there are more destinations available 10–100 km away due to urban sprawl. While these causal factors are correlated with density, urban form can substantially affect travel energy use even within a given density bracket. We also show that urban form predominantly influences how and where people travel, rather than how often and how long. These results outline pathways for cities and communities to reduce the environmental impact of travel while increasing access to relevant destinations.