Abstract A stated-preference approach is used to elicit the attitudes of the general public towards coyote conservation. The payment vehicle is presented in a way that explicitly prompts respondents to adopt a citizen perspective, rather than a consumer perspective, when responding to the survey. To deal with the large numbers of zero responses, a Box-Cox Double-Hurdle specification is used to model separately individuals' choices about whether to support conservation or not and their choice about the level of support. The results show that simpler analyses that do not account explicitly for these two different decisions would lead to misleading conclusions in the study of nuisance wildlife. The study uses primary data from Prince Edward Island (Canada).