Carnivore conservation and management are global research priorities focused on reversing population declines of imperiled species and identifying more effective and humane management of generalist carnivores with thriving populations. Nonlethal methods to mitigate conflict are increasingly used to advance conservation objectives; however, there is limited knowledge about the effectiveness of many nonlethal methods. We tested a nonlethal tool (fladry), that serves as a barrier to deter wolves Canis lupus and coyotes Canis latrans, for its efficacy at preventing coyotes from using prairie dog Cynomys ludovicianus colonies, the primary prey for critically endangered black-footed ferrets Mustela nigripes. We used camera trap data and an occupancy approach to evaluate the tool’s efficacy. We measured coyote response to fladry at both a coarse monthly scale (via use, attraction and avoidance probabilities) and a fine scale (via daily activity). Overall, use of areas inside exclosures declined by 60% after 60 days of fladry application and coyotes avoided some previously used areas both within and outside exclosures. Interestingly, coyotes were attracted to previously unused areas surrounding exclosures and increased activity around the periphery of fladry exclosures by 170% immediately after fladry installation, suggesting coyotes actively explored these areas and may have responded to fladry in a way that is counterintuitive to management expectations. Occupancy models provided more robust evaluation of fladry and revealed important behavioral responses relative to other common evaluation techniques (i.e. time until first detected crossing). Our results have implications for future development and evaluation of nonlethal tools for carnivore conservation and management globally.