1. Habitat loss, fragmentation and degradation are key threats to the long-term persistence of carnivores, which are also susceptible to direct persecution by people. Integrating natural and social science methods to examine how habitat configuration/quality and human-predator relations may interact in space and time to effect carnivore populations existing within human-dominated landscapes will help prioritise conservation investment and action effectively.2. We propose a socio-ecological modelling framework to evaluate drivers of carnivore decline in landscapes where predators and people coexist. By collecting social and ecological data at the same spatial scale, candidate models can be used to quantify and tease apart the relative importance of different threats.3. We apply our methodological framework to an empirical case study, the threatened guiña (Leopardus guigna) in the temperate forest ecoregion of southern Chile, to illustrate its use. The existing literature suggests that the species is declining due to habitat loss, fragmentation and persecution in response to livestock predation. Data used in modelling were derived fromfour seasons of camera-trap surveys, remote-sensed images and household questionnaires.4. Occupancy dynamics were explained by habitat configuration/quality covariates rather than by human-predator relations. Guiñas can tolerate a high degree of habitat loss (>80% within a home range). They are primarily impacted by fragmentation and land subdivision (larger farms being divided into smaller ones). Ten percent of surveyed farmers (N=233) reported illegally killing the species over the past decade.5. Synthesis and applications. By integrating ecological and social data into a single modelling framework, our study demonstrates the value of an interdisciplinary approach to assessing the potential threats to a carnivore. It has allowed us to tease apart effectively the relative importance of different potential extinction pressures, make informed conservation recommendations and prioritise where future interventions should be targeted. Specifically for the guiña, we have identified that human-dominated landscapes with large intensive farms can be of conservation value, as long as an appropriate network of habitat patches are maintained within the matrix. Conservation efforts to secure the long-term persistence of the species should focus on reducing habitat fragmentation, rather than human persecution in our study system.