It has been argued that respondents in contingent valuation (CV) surveys, asked to value complex environmental amenities, will state willingness to pay (WTP) independently of the scope of the project. Such insensitivity to scope would be at odds with rational choice, and could therefore imply that CV is not a theoretically valid method for biodiversity valuation. The scope test in the present CV study was applied to endangered species preservation. Respondents were split in four sub-samples facing different scopes of endangered species preservation. The design allowed for both external and internal scope tests. Furthermore, the tests were split according to elicitation format. Of four external tests of insensitivity to scope, one was rejected, two gave mixed results, depending on either the type of test or elicitation format, and for the last one the null hypothesis could not be rejected. Of five internal tests, insensitivity to scope was rejected in three cases, one test gave mixed results, and one could not be rejected. Survey design features of the CV study, especially an unfamiliar sub-group of endangered species, could explain the apparent insensitivity to scope observed.