We report on quantum chemical calculations at the DFT (BP86/TZP) and ab initio (CCSD(T)/III+) levels of the title compounds. The geometries, vibrational spectra, heats of formation, and homolytic and heterolytic bond dissociation energies are given. The calculated bond length of Cu-CN is in reasonable agreement with experiment. The theoretical geometries for CuNC and the other group 11 cyanides and isocyanicles which have not been measured as isolated species provide a good estimate for the exact values. The theoretical bond dissociation energies and heats of formation should be accurate with an error limit of +/-5 kcal/mol. The calculation of the vibrational spectra shows that the C-N stretching mode of the cyanides, which lies between 2170 and 2180 cm(-1), is IR inactive. The omega(1)(C-N) vibrations of the isocyanicles are shifted by similar to100 cm(-1) to lower wavenumbers. They are predicted to have a very large IR intensity. The nature of the metal-ligand interactions was investigated with the help of an energy partitioning analysis in two different ways using the charged fragments TM+ + CN- (TM = transition metal) and the neutral fragments TM. + CN. as bonding partners. The calculations suggest that covalent interactions are the driving force for the formation of the TM-CN and TM-NC bonds, but the finally formed bonds are better described in terms of interactions between TM+ and CN-, which have between 73% and 80% electrostatic character. The contribution of the pi bonding is rather small. The lower energy of the metal cyanides than that of the isocyanicles comes from the stronger electrostatic interaction between the more diffuse electron density at the carbon atom of the cyano ligand and the positively charged nucleus of the metal.