The chemisorption of an acetylene (HCCH), a vinylidene (CCH2), and an ethylidyne (CCH3) on metal surfaces, especially Pt(lll), is analyzed in some detail, with an emphasis on the electronic rearrangements ensuing. The bonding is described in terms of semilocalized states, in turn obtained via a deconvolution of the total density of states into fragment orbitals of the hydrocarbon and the surface. The geometrical choices made by the various fragments on surfaces are analyzed by simple perturbation theory. It is found that in general the bonding within both the hydrocarbon fragment and the surface is dramatically weakened, as indicated by the changes in overlap populations that result. The more surface atoms are involved in anchoring the adsorbate, the weaker the bonding within the surface becomes. A nondissociative chemisorption is the result of a compromise and is operative when the price of the binding energy is evenly distributed over the loss of bonding within the adsorbate and the surface. Surface reconstruction and dissociative chemisorption are to be expected as a consequence of the two extreme situations that may occur:A filling of metal-metal antibonding states would lead to surface reconstruction, whereas excessive population of carbon-carbon antibonding orbitals would drive a dissociative chemisorption. The two patterns are illustrated by a would-be 4-fold adsorption of C2H2 on Pt(lll) and the chemisorption of C2H2 on Fe(100), respectively. In the latter it is shown that the entire carbon-carbon bonding energy is passed to the metal-carbon bonds, in a kind of bonding transfer. We find a special role for the bulk-centered states, that of a reservoir which can be alternatively filled or emptied depending on the metal atoms constituting the surface. Throughout the paper a comparison is made between the binding of the fragments to the metallic surface and to discrete transition-metal fragments in organometallic chemistry.Chemistry has been enriched in the last 2 decades by an exponentially growing literature dealing with small molecules chemisorbed on metal surfaces. The reason for this expansion of our knowledge lies in a combination of three factors: (i) an advancing technology which has allowed the development of highly sophisticated spectroscopic tools relevant to surface studies,1 (ii) an increased interest in many catalysis-related phenomena,2 34and (iii)(1) (a) Ertl, G.;