A recent theory of ionization of carboxylic acids divided the process of ionization into two steps and attributed the main importance to the electrostatic potential of the acid molecule. The origin of the acidity was thus seen in the high energy of the acid molecule and not in the stabilization of the anion by resonance. In this paper, the theory was revised on the basis of further calculations at an MP2/6-31++G(2d,p) level which followed in two steps the reverse process, protonation of the formate anion. The results were compared to those of the methanolate anion as reference. A contradictory conclusion was obtained: the reason for the acidity is in the electrostatic potential of the anion. As a model compound without resonance, 2,2,2-trifluoroethanol was investigated in the same way. The results were qualitatively similar, although any resonance in the anion is not possible. One can conclude that the acidity of carboxylic acids is due to the low energy of their anions; this follows unambiguously from the isodesmic reactions on the basis of either the experimental enthalpies of formation or the calculated energies. It is more difficult to decide whether this low energy is caused by resonance, because any model of the carboxylate anion without resonance is necessarily imperfect and the whole concept is not exactly defined. Several such models were reexamined and improved by separating the effect operative in the anion from those in the neutral acid molecule. While the electrostatic model did not allow any significant conclusion, two models based on VB calculation and on correlation analysis, respectively, furnished qualitatively concordant results: resonance in the anion is diminished by the resonance in the acid molecule, and both together are responsible for less than one-half of the acidity enhancement as compared to the acidity of alcohols. The ratio is reversed in water solution: resonance is then the more important factor responsible for some two-thirds of the enhanced acidity.