Conserving endangered species sometimes involves killing their predators. In the case of Nearctic wolves (Canis lupus), rarely are lethal control measures examined for ancillary effects on predator behaviour or community responses in a before-after design. We examined wolf relative abundance and spatial distribution in a northwestern boreal forest landscape for three years before and after the onset of wolf culling intended to conserve threatened woodland caribou (Rangifer tarandus caribou). We hypothesized that wolf occurrence would increase with density of anthropogenic features created by landscape development before the cull, but that wolves would avoid anthropogenic features after the cull due to associated mortality risk. We used generalized linear models in an information-theoretic framework to weigh evidence for our hypotheses. Post-control, independent wolf detections decreased to 24 % of pre-cull numbers, but wolves maintained 75 % of their distribution. Pre-control, wolves were positively associated with linear features, presumably for hunting efficiency, but post-cull wolves were negatively associated with these features. Thus, wolf control caused not only a numerical reduction of wolf numbers, but also a functional change in wolf behaviour that could further reduce predation pressure on caribou. However, post cull wolf occurrence was more strongly associated with anthropogenic block features which provide forage for alternate prey, potentially subsidizing their fast recovery. Conservation actions involving predator mortality alter landscape-scale distributions and behaviors of surviving predators, with potential indirect effects for the mammal community.