Managers of nuisance wildlife have to rely largely on using lethal methods until such time as nonlethal techniques, such as fertility control, become universally available for a wide range of species. Unfortunately, use of lethal tools has met with opposition from animal welfare and animal rights proponents. Although research has addressed some of the more tractable welfare concerns (e.g., making traps more humane), less tractable ethical issues associated with the justification of killing wildlife remain unresolved. Monistic welfare models or rights-based models have been proposed as ways of addressing these issues, but those that concentrate on the cognitive and conative capabilities of individual animals fail to resolve the ecological and social complexities involved in management of nuisance wildlife. Solutions need to recognize and accept the diversity of values (i.e., within a pluralistic strategy) as well as the uncertainty inherent in many of the systems being managed. Thus, when uncertainty is high in managing wildlife–resource systems, we propose the only ethically defensible action is to apply a knowledge-based ethic that ensures future actions will be carried out with increased understanding. Such an ethic can be made functional within an adaptive management framework that has, as its first tenet, the need to learn and reduce uncertainty. Failure to maximize learning in the presence of uncertainty has the potential to result in increased opposition to even soundly justified operations to manage nuisance wildlife.