Although poaching (illegal killing) is an important cause of death for large carnivores globally, the effect of lethal management policies on poaching is unknown for many populations. Two opposing hypotheses have been proposed: liberalizing killing may decrease poaching incidence ('tolerance hunting') or increase it ('facilitated poaching'). for gray wolves in Wisconsin, USA, we evaluated how five causes of death and disappearances of monitored, adult wolves were influenced by policy changes. We found slight decreases in reported wolf poaching hazard and incidence during six liberalized killing periods, but that was outweighed by larger increases in hazard and incidence of disappearance. Although the observed increase in the hazard of disappearance cannot be definitively shown to have been caused by an increase in cryptic poaching, we discuss two additional independent lines of evidence making this the most likely explanation for changing incidence among n = 513 wolves' deaths or disappearances during 12 replicated changes in policy. Support for the facilitated poaching hypothesis suggests the increase (11-34%) in disappearances reflects that poachers killed more wolves and concealed more evidence when the government relaxed protections for endangered wolves. We propose a refinement of the hypothesis of 'facilitated poaching' that narrows the cognitive and behavioral mechanisms underlying wolf-killing. Globally, loss of large predators has contributed to simplification of trophic structures, lower biodiversity and degradation of ecosystem functions 1-3. It is widely acknowledged that humans are responsible for more large carnivore deaths than any other cause 4 , although the scientific debate about the sustainability of this killing remains far from settled for many large carnivore populations 3,5,6. Many policies for large carnivores focus on limiting or regulating human-caused mortality, and many management decisions rely on estimates of human-caused mortality and on understanding the policy effects on such mortality. Therefore, biased estimates of human-caused mortality patterns can undermine policy goals and evaluations (e.g., recolonization by endangered species, restoring ecosystem processes), and harm carnivore population recovery or stability 5-7. More broadly, accurate estimates of policy effects on illegal killing ('poaching' hereafter) of wild animals can improve enforcement of laws for nature protection and adherence to national and international treaties relating to the protection or the restoration of endangered species, ecosystems, biodiversity, and interdiction of unregulated wildlife trade. Of all direct killing by humans, poaching is the primary cause of death in many carnivore populations 5,8-11 , slowing population growth 12-14 or hindering recolonization of historic range 8,10. Poaching is extremely difficult to detect, measure and prevent. Given its illegal nature, poachers often conceal evidence from management authorities tasked with monitoring marked animals. When authorities find the bod...