Quantum chemical techniques today are indispensable for the detailed mechanistic understanding of catalytic reactions. The development of modern density functional theory approaches combined with the enormous growth in computer power have made it possible to treat quite large systems at a reasonable level of accuracy. Accordingly, quantum chemistry has been applied extensively to a wide variety of catalytic systems. A huge number of problems have been solved successfully, and vast amounts of chemical insights have been gained. In this Account, we summarize some of our recent work in this field. A number of examples concerned with transition metal-catalyzed reactions are selected, with emphasis on reactions with various kinds of selectivities. The discussed cases are (1) copper-catalyzed C-H bond amidation of indoles, (2) iridium-catalyzed C(sp(3))-H borylation of chlorosilanes, (3) vanadium-catalyzed Meyer-Schuster rearrangement and its combination with aldol- and Mannich-type additions, (4) palladium-catalyzed propargylic substitution with phosphorus nucleophiles, (5) rhodium-catalyzed 1:2 coupling of aldehydes and allenes, and finally (6) copper-catalyzed coupling of nitrones and alkynes to produce β-lactams (Kinugasa reaction). First, the methodology adopted in these studies is presented briefly. The electronic structure method in the great majority of these kinds of mechanistic investigations has for the last two decades been based on density functional theory. In the cases discussed here, mainly the B3LYP functional has been employed in conjunction with Grimme's empirical dispersion correction, which has been shown to improve the calculated energies significantly. The effect of the surrounding solvent is described by implicit solvation techniques, and the thermochemical corrections are included using the rigid-rotor harmonic oscillator approximation. The reviewed examples are chosen to illustrate the usefulness and versatility of the adopted methodology in solving complex problems and proposing new detailed reaction mechanisms that rationalize the experimental findings. For each of the considered reactions, a consistent mechanism is presented, the experimentally observed selectivities are reproduced, and their sources are identified. Reproducing selectivities requires high accuracy in computing relative transition state energies. As demonstrated by the results summarized in this Account, this accuracy is possible with the use of the presented methodology, benefiting of course from a large extent of cancellation of systematic errors. It is argued that as the employed models become larger, the number of rotamers and isomers that have to be considered for every stationary point increases and a careful assessment of their energies is therefore necessary in order to ensure that the lowest energy conformation is located. This issue constitutes a bottleneck of the investigation in some cases and is particularly important when analyzing selectivities, since small energy differences need to be reproduced.