I discuss the prospects of perspectival realism for resolving the problem of incompatible models or theories in scientific practice. My diagnosis is that the perspectivist can secure the "realism' in her position only by employing suitable relations between the models. It is such relations that do the work, not the general philosophical claim about the perspectival nature of knowledge claims. But appeal to such relations has also been the preferred strategy of scientific realist approaches to the problem. With respect to the problem of incompatible models, then, it is not clear that perspectivism has a clear advantage—even though the issue was among the motivating problems for proposing this view. I suggest, however, that the required inter-model relations (typically limit relations) embody a narrower notion of perspective, a notion that is explicitly part of the models themselves and that is essential in maintaining a realist view.