AbstractThe classical concept of molecular structure (namely, a set of atoms with a well‐defined geometrical arrangement in space) originated well before the advent of quantum theory, and, contrary to widespread views, these two conceptual schemes (classical molecular structure and quantum theory) are not easy to reconcile. We begin with a preliminary discussion of molecular structure in terms of space and time correlations in the framework of quantum theory. Then we summarize the salient features of some critical views raised‐during recent years (essentially by Woolley) concerning the concept of molecular structure. The essential claim of this viewpoint is that “molecular structure” is not an intrinsic attribute of an isolated molecule, and appears rather as an effect of the environment (other molecules? vacuum electromagnetic field, according to quantum electrodynamics?). Afterwards, we discuss these issues in detail: we show that it is necessary to distinguish between a “quantum” or “potential” structure (which may be deduced in a straightforward way from the quantum treatment of the molecule), and the “classical” molecular structure (according to which the molecule is considered as a set of rotating and vibrating material points, corresponding to the atoms); we briefly discuss some aspects of the Born–Oppenheimer approximation. From our discussion, we deduce that the “classical structure” concept is indeed non‐trivial from a purely quantum‐mechanical point of view; in actual fact, it is related to the problem of the “classical limit” of quantum mechanics, and it appears that, at the present time, none of these classical concepts and behaviours is convincingly derivable from a strict quantum mechanical basis. Finally, we consider these same problems from the point of view of stochastic electrodynamics (SED), which is one of the stochastic models for microphysics proposed so far as a possible alternative or extension of usual quantum theory. In the framework of such a theory, the “classical limit” and, in particular, the “classical molecular structure” no longer poses a problem, but aside from these satisfactory qualitative aspects, several quantitative failures of SED prevent us from accepting it, at least in its present state, as a genuine physical alternative to quantum theory. We therefore conclude that “classical molecular structure”, and classical concepts in general, remain an open problem in