The acidities of 18 5,6-substituted uracils have been numerically estimated as pK(a) values in terms of three theoretical models. The first scheme includes the calculation of the gas-phase acidity of uracil with the G3MP2B3 method and taking into account the solvent effect using the polarizable continuum approximation PCM(SMD)-TPSS/aug-cc-pVTZ. The second model is one step and implies calculation of the free Gibbs energies of the hydrate complex of uracil (and its anion) with 5 water molecules by the TPSS/aug-cc-pVTZ method. This model accounts for the solvent effect corresponding to both specific and nonspecific solvation. The third scheme required high time and computational resources and includes the strong features of the two previous schemes. Here, the theoretical estimation of pK(a) is performed by the CBS-QB3 composite method. As in the second approach, both specific (as pentahydrate) and nonspecific solvent effects are determined. We have analyzed the advantages and model restrictions of the considered schemes for the pK(a) calculations. All models have systematic errors, which have been corrected with the linear empirical regression relations. In the presented model, the absolute mean deviations of the pK(a) values of uracils dissociating via the N1-H bonds diminish to 0.25, 0.28, and 0.23 pK(a) units (respectively, for I, II, and III model's), which corresponds to similar to 0.3 kcal/mol on the energy scale. The applicability of our computational schemes to uracils dissociating via N3-H, O-H (orotic acids) and C-H bonds is discussed.