Both KNI-10033 and KNI-10075 are high affinity preclinical HIV-1 protease (PR) inhibitors with affinities in the picomolar range. In this work, the molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) method has been used to investigate the potency of these two HIV-1 PR inhibitors against the wild-type and mutated proteases assuming that potency correlates with the affinity of the drugs for the target protein. The decomposition of the binding free energy reveals the origin of binding affinities or mutation-induced affinity changes. Our calculations indicate that the mutation I50V causes drug resistance against both inhibitors. On the other hand, we predict that the mutant I84V causes drug resistance against KNI-10075 while KNI-10033 is more potent against the I84V mutant compared to wild-type protease. Drug resistance arises mainly from unfavorable shifts in van der Waals interactions and configurational entropy. The latter indicates that neglecting changes in configurational entropy in the computation of relative binding affinities as often done is not appropriate in general. For the bound complex PRI50V-KNI-10075, an increased polar solvation free energy also contributes to the drug resistance. The importance of polar solvation free energies is revealed when interactions governing the binding of KNI-10033 or KNI-10075 to the wild-type protease are compared to the inhibitors darunavir or GRL-06579A. Although the contributions from intermolecular electrostatic and van der Waals interactions as well as the nonpolar component of the solvation free energy are more favorable for PR KNI-10033 or PR-KNI-10075 compared to PR-DRV or PR GRL-06579A, both KNI-10033 and KNI-10075 show a similar affinity as darunavir and a lower binding affinity relative to GRL-06579A. This is because of the polar solvation free energy which is less unfavorable for darunavir or GRL-06579A relative to KNI-10033 or KNI-10075. The importance of the polar solvation as revealed here highlights that structural inspection alone is not sufficient for identifying the key contributions to binding affinities and affinity changes for the design of drugs but that solvation effects must be taken into account. A detailed understanding of the molecular forces governing binding and drug resistance might assist in the design of new inhibitors against HIV-1 PR variants that are resistant against current drugs.