The results of ab initio calculations are reported for prototypical high-valent, alkylidene complexes. Stationary points on each potential energy surface are characterized and compared to experimental information where available; as long as a suitably flexible valence basis set is used, good agreement between theoretically calculated and experimentally determined geometries is obtained. The complexes of interest include group IVB (Ti, Zr and Hf) and group VB (Nb and Ta) alkylidenes with hydride ligands as well as models for the four-coordinate, olefin metathesis catalysts (Mo-, W-, and Re-alkylidenes) which have been recently synthesized and characterized. In light of the fact that much of the discussion concerning the reactivity of transition-metal car bene complexes has been presented in terms of the resonance contributors derived from rearranging the electrons in the M-e u and 1r orbitals, the minima obtained from the first portion of the study are then subjected to a further procedure to calculate these contributions. Resonance structures in which the carbon is the negative end of the M-e bond (i.e., nucleophilic resonance structures) contribute 50% to the ground-state wave function of these complexes. Those in which the carbon is formally neutral account for much of the remainder ( 45% ). Only 5% is comprised of electrophilic resonance structures, i.e., those in which the carbon is the positive end of the M-C bond. Furthermore, the metal-carbon double bond is predominantly comprised of five resonance structures. Four of these resonance structures correspond to models of carbene bonding which have been discussed previously in the literature. The other resonance structure, which contributes roughly 33% to the ground-state wave function, has hitherto not been considered when examining the chemical reactivity of carbenes. This large resonance contributor can be described as arising from a dative carbon-to-metal u bond plus a covalent M-e 1r bond.