Understanding the interplay between exploitation and natural mortality is essential to guiding sustainable conservation of wildlife. Exploitation of carnivores by humans has long been thought to result in compensatory reductions of natural mortality among survivors. If rates of human exploitation exceed natural mortality, however, such actions will ‘add’ to overall mortality and could imperil the sustainability of such actions. We applied competing risk analyses to ⩾16 years of data for heavily harvested and semi-protected cougar populations in Utah to test the additive and compensatory mortality hypotheses, while accounting for parameter uncertainty. We additionally tested for presence of the two primary mechanisms by which compensatory mortality can arise: density dependence and individual heterogeneity in mortality risks...