We present herein an appraisal on the performance of the generalized Born (GB) model in estimating the solvation energies of small molecules and pK(a) shifts of dicarboxylic acids. The quality of the solvation energy results obtained with the GB model was exceedingly good as already reported in the literature but the pK(a) shift estimates fell short of expectations. Analysis of the problem on a simple prototype system revealed that with the GB model, the estimates of the two components, viz. the shielding and the self-energy terms, to be somewhat in error. These errors compensate each other in the calculation of solvation energies but affect the intramolecular interaction energies and hence pK shifts differently. We examine here the feasibility of introducing modifications to the GB model for a simultaneous evaluation of both solvation and intramolecular interaction energies.