In this paper I attempt to show, against certain versions of trope theory, that properties with analyzable particularity cannot be merely exactly similar: such properties are either particularized properties (tropes) that are dissimilar to every any other trope, or else universalized properties (universals). I argue that each of the most viable standard and nonstandard particularizers that can be employed to secure the numerical difference between exactly similar properties can only succeed in grounding the particularity of properties, that is, in having (...) properties be tropes, at the expense of ruling out the possibility of their exact similarity. Here are the four nonstandard particularizers that I examine: the genealogy of a property, the history of a property, the causal effects of a property, and the duration of a property. And here are the two standard particularizers that I examine: the bearer of a property, by which I mean either a bare particular or a spatiotemporal location, and the property itself, by which I mean that the property is self-particularized. In my concluding remarks, I explain that the only remaining hope for preserving the possibility of exactly similar tropes is regarding properties as primitively particular, and that this must mean not that properties are selfparticularized but that they are particularized due to nothing. I close by arguing that this may not help trope theory after all. (shrink)