The reaction energy profiles of the benzoin condensation from three aldehydes catalyzed by imidazol-2-ylidene, triazol-3-ylidene, and thiazol-2-ylidene have been investigated computationally. The barriers for all steps of all investigated reactions have been found to be low enough to indicate the viability of the mechanism proposed by Breslow in the 1950s. The most remarkable difference in the catalytic cycles has been the increased stability of the Breslow intermediate in case of thiazol-2-ylidene (by ca. 10 kcal/mol) compared to the other two carbenes, which results in lower energy for the coupling of the second aldehyde molecule, thus, increasing the reversibility of the reaction. Since the analogous transketolase reaction, being involved in the carbohydrate metabolism of many organisms, requires an initial decoupling-a reverse benzoin condensation-this difference provides a reasonable explanation for the presence of a thiazolium ring in thiamine instead of the otherwise generally more available imidazole derivatives. The "resting intermediate" found by Berkessel and co-workers for a triazole-based catalyst was found more stable than the Breslow intermediate for all of the systems investigated. The (gas phase) proton affinities of several carbenes were compared, the relative trends being in agreement with the available (in aqueous solution) data. The hydrolytic ring-opening reaction of the thiazole-based carbene was shown to be different from that of imidazole-2-ylidenes.