AbstractUrban trees are increasingly used by cities for cooling and climate adaptation. However, efforts to increase tree cover across cities have neglected to account for the trees' health and function, which are known to control their associated environmental benefits but have been difficult to assess at scales relevant for management. Here, we use remotely sensed, high resolution canopy temperature as a proxy for tree health and function and evaluate its relation to the built environment across Minneapolis‐St. Paul (MSP) using machine learning analyses. We develop a new index that incorporates information on urban trees' health and function, in addition to their presence. This index, when applied across MSP, suggests that canopy benefits may not be distributed equally even in neighborhoods with similar canopy cover. Furthermore, accounting for tree health and function can yield more effective and equitable benefits by guiding the location and magnitude of intervention for urban tree management.