Reintroducing threatened mammals into the broader landscape outside fenced reserves requires effective control of feral predators and knowledge of predator density. Most robust methods of density estimation require the identification of individual animals, possible for cats but not foxes. This project is developing new statistical methods to generate robust estimates based on camera trap data. The models use knowledge of space use derived from GPS collars to model encounter rates without the need for individual ID. These estimates will be used to assess the effectiveness of predator control measures (e.g., baiting and trapping) and control strategies (e.g., timing, frequency and extent). They will ultimately be used to identify thresholds of predator density that enable the safe release of threatened mammals.