A survey of the specificities of the electronic structure of transition metal carbonyls is presented, with emphasis on the theoretical complication inherent to the description of this class of molecules. The difficulties originate from the nature of the metal‐carbonyl bonding, which is governed by the electronic configuration of the metal atom. The distribution of the electronic density may vary significantly from first‐row to third‐row transition metals. The presence of several metal centers or other ligands than CO perturbs the well‐balanced repartition driven by the donor‐acceptor capabilities of the metal and the CO groups in “classic” saturated M(CO)
n
complexes. Quantum chemistry needs highly correlated methods, which are sufficiently flexible to model this electronic flexibility that is at the origin of the chemical richness of transition metal carbonyls. A short review of the appropriate methods, with their advantages and drawbacks, is given in the introductory section.
Several classes of transition metal carbonyls are described in the case studies section. The theoretical analysis of electronic ground and excited states properties of neutral and charge M(CO)
n
saturated first‐, second‐, and third‐row transition metal carbonyls is illustrated for M(CO)
6
(M = Cr, Mo, W), M(CO)
5
(M = Fe, Ru, Os), and M(CO)
4
(M = Ni, Pd, Pt). More predictive theoretical work concerns either the unsaturated species, generated by photofragmentation, which are characterized by nearly degenerate electronic ground states such as MCO, M(CO)
2
, and M(CO)
3
, or transition metal carbonyls with coordination numbers greater than six. Density functional theory (DFT) methods as well as ab initio approaches are necessary to study these complicated electronic problems. Here, the field of polynuclear transition metal carbonyls is limited to binuclear complexes, studied extensively by DFT owing to the size of the molecules characterized by numerous geometrical structures with terminal or bridged CO and the possibility for metal–metal multiple bonds. This is illustrated by a detailed study of binuclear vanadium carbonyls V
2
(CO)
n
(
n
= 9–12) coexisting in singlet as well as in triplet electronic configurations. Even if binuclear transition metal carbonyls are a source of interesting photochemistry, extensively investigated experimentally, only a few theoretical studies are devoted to this aspect. The world of mixed ligand transition metal carbonyls is vast and the number of problems to be solved is unlimited. One important part concerns the electronic ground state reactivity that can be studied by routine calculations based on the determination of reaction pathway energetics, transition states, and intermediates structures. The discovery of new catalysts is another facet illustrated by the study of the bonding capabilities of various ligands, which are isolobal analogs of CO. In the same manner, DFT calculations allow the prediction of new possible structures isoelectronic to well‐known homoleptic binuclear iron and manganese carbonyls. The richness of the photophysics and photochemistry of mixed ligand transition metal carbonyls makes them an inexhaustible subject for investigation including the assignment of featureless absorption spectra, the modeling of elementary steps that contribute to the observed photoreactivity, or the determination of the complete structure of electronic excited states. Transition metal carbonyls possessing low‐lying metal‐to‐ligand‐charge‐transfer states are fascinating in the diversity of processes generated by light absorption in the UV‐vis energy domain such as ultrafast dissociation of CO (∼100 fs), luminescence, electron transfer, or isomerization of selective ligands.
