Zinc is an endogenous part of many proteins. Cadmium is a toxic metal, and replacing zinc in proteins is one of the reasons for its toxicity. Reliable estimates of the interaction energies between the ions and their ligands in proteins are desired, but difficult to obtain. Using small molecular models for Zn and Cd ligands in proteins, Emma Ahlstrand et al. on page 2554 calculate interaction energy curves with MP2 and DFT functionals, and compare them to CCSD(T) reference values. The results call for caution in applying DFT methods to metalloprotein model complexes, in particular at ion–ligand separations that are longer than the equilibrium distances.