I argue that molecules may not have structure in isolation. I support this by investigating how quantum models identify structure for isolated molecules. Specifically, I distinguish between two sets of models: those that identify structure in isolation and those that do not. The former identify structure because they presuppose structural information about the target system via the Born-Oppenheimer approximation. However, it is an idealisation to assume structure in isolation because there is no empirical evidence of this. In fact, whenever structure is empirically examined it is always partially determined by factors that are absent in isolation. Together with the growing empirical evidence that isolated molecules behave in non-classical ways, this shows that the quantum models that do not identify structure are more faithful representations of isolated molecules.Preprint of chapter In upcoming Book: Philosophical Perspectives on Quantum Chemistry, ed. by Olimpia Lombardi,