There is a growing awareness of problems with the estimation of the ecological tolerances of species through correlative modeling approaches. These problems have led some investigators to argue for abandoning terms such as 'ecological niche model' and 'environmental niche model' in favor of the ostensibly more value-neutral 'species distribution model', as the models are thought to frequently be poor estimators of the niche. Here, I argue that most applications to which these models are put require the assumption that they do estimate the niche, however imperfectly, and that obscuring this inescapable and potentially flawed assumption in the terminology may only serve to hinder the development of the field.