This essay examines the underdetermination problem that plagues structuralist approaches to spacetime theories, with special emphasis placed on the epistemic brands of structuralism, whether of the scientific realist variety or not. Recent non-realist structuralist accounts, by Friedman and van Fraassen, have touted the fact that different structures can accommodate the same evidence as a virtue vis-à-vis their realist counterparts; but, as will be argued, these claims gain little traction against a properly constructed liberal version of epistemic structural realism. Overall, a broad construal of spacetime theories along epistemic structural realist lines will be defended which draws upon both Friedman's earlier work and the convergence of approximate structure over theory change, but which also challenges various claims of the ontic structural realists.The contemporary revival of structuralism in the philosophy of science can be traced to a host of early twentieth century structuralist approaches, principally, the philosophical speculation on the epistemology and ontology of mathematical physics put forward by Poincaré, Eddington, Weyl, Russell, and Cassirer, to name just a few. Given the structuralist's fixation on the mathematical and/or empirical structure of theories, as opposed to, say, the underlying theoretical entities alleged to bring about and sustain those structures, it is therefore not surprising that the underdetermination of structure quickly became a serious obstacle for the new program; e.g., Newman's critique of Russell (see footnote 10). That is, since more than one structure is liable to be consistent with the same evidence, a theory's structure is as underdetermined as its ontology (for, as is well known, several different sets of unobservable entities can be reconciled with the same observation basis). This essay will examine the current