AbstractThe idea that a single molecule has a shape—that is, a longlived nearly fixed (relative) arrangement of nuclei in space—has been shown by Woolley to be a classical idea imposed on the quantum mechanical picture of matter. In this report the argument that the Born–Oppenheimer approximation is semi‐classical is reviewed, and some pioneering computations free of the Born–Oppenheimer approximation are discussed. The problem of reconciling the results of these computations with the powerful and successful molecular structure theory is described, and a mode of computation is proposed which strikes a compromise between the sometimes inadequate Born–Oppenheimer Hamiltonian and the complicated full molecular