Spatial heterogeneity in the local densities of terrestrial carnivores is driven by multiple interacting biotic and abiotic factors. Space-use patterns of large carnivores reflect the competing demands of resource selection (e.g., exploitation of habitats with abundant prey) and minimization of risks arising from human interactions. Estimating the relative strength of these drivers is essential to understand spatial variation in densities of large carnivores and there are still key knowledge gaps for many large carnivore populations. To better understand the relative roles of environmental and human drivers of spatial variation in tiger ( Panthera tigris ) densities, we surveyed a 3000 km 2 landscape in North India using camera trap data. Over two years, we photo-captured 92 unique adult tigers. Associating spatial covariates with patterns of detection allowed us to test hypotheses about the relative influence of prey abundance, habitat structure and extent, and proximity to habitat edges on spatial variation in tiger densities across a gradient of anthropogenic disturbance. We documented extensive variation in tiger density within and across management units and protected areas. Spatial variation in prey abundance and proximity to grassland habitats, rather than human use (e.g. extent of human-dominated edge habitat and protection status), explained most of the spatial variation in tiger density in two of the five surveyed sites. The region’s largest tiger population occurred in a multi-use forest beyond protected area boundaries, where wild ungulates were abundant. Our results suggest that tigers can occur at high densities in areas with extensive human use, provided sufficiently high prey densities, and tracts of refuge habitats (eg. areas with dense vegetation with low human use). We argue that tiger conservation portfolio can be expanded across multi-use landscapes with a focus on areas that are adaptively managed as “zones of coexistence” and “refuge habitats”. Advancing this conservation strategy is contingent on greatly strengthening systems to effectively and equitably redress human–wildlife conflict and leveraging existing policies to strengthen local participation in conservation planning and forest stewardship. Our insights into the environmental drivers of spatial heterogeneity in tiger populations can inform both local management and guide to species recovery in working landscapes.