Mechanisms of the gas-phase acyl group transfers, Cl- + R(X=Y)Cl, involving various acyl functional groups, >X=Y with X = C, S, or P and Y = O or S, are investigated theoretically at the MP2/6-31+G* and B3LYP/6-31+G* levels (additionally with extended basis sets of B3LYP/6-311+G(3df.2p)), and the effects of solvent (epsilon = 78.5) are calculated with the SCIPCM model at the isodensity level of 0.0003 au. The tetrahedral adducts formed in the carbonyl (RC=O) and thiocarbonyl (RC=S) group transfers are either transition states (double-well PES) or intermediates (single- or triple-well PES) depending on R, a stronger electron acceptor R favoring the intermediate. However, all of the sulfonyl (RSO2) and phosphoryl ((RO)(2)P= O) transfers proceed with trigonal bipyramid (TBP)-type transition states, in contrast to the stepwise mechanism through TBP-type intermediates for the sulfinyl (RS=O) land sulfonyl transfers between F-) transfers. The most important factor determining whether an adduct in an acyl-group-transfer reaction is the transition state or intermediate is the energy gap between the pi*(X-Y) and sigma*(X-LG) orbitals. The possibility of reacting through an intermediate is greater for lower pi*(X-Y) and higher sigma*(X-LG) levels. The backside sigma -attack pathway is favored over the pi -attack pathway only when a low-lying sigma*(X-LG) orbital, preferably below the pi*(X-Y) level, is available. In general, the results are in good agreement with those of experiments. The solvent effect elevates the barrier height almost uniformly so that the relative orders of gas-phase activation barriers between different R groups are maintained in solution.