Some of drug-resistant mutants of HIV-1 protease (PR), such as a clinically-relevant drugresistant PR mutant (Flap+ (I54V) ) containing L10I, G48V, I54V and V82A mutations, produce significant changes in the balance between entropy and enthalpy of the drug-PR interactions, compared to the wild-type (WT) PR. Here, to gain a comprehensive understanding of the entropy-enthalpy compensation effects, we compared nuclear magnetic resonance (NMR), fluorescence spectroscopy and isothermal titration calorimetry (ITC) data of a WT PR with Flap+ (I54V) and related mutants: (1) Flap+ (I54V) ; (2) Flap+ (I54A) which evolves from Flap+ (I54V) in the continued presence of inhibitor yet does not exhibit entropy-enthalpy compensation; and (3) Flap+ (I54) , a control mutant that contains only L10I, G48V and V82A mutations. Our data indicate that WT and Flap+ (I54A) show enthalpy-driven inhibitor-interaction, while Flap+ (I54) and Flap+ (I54V) exhibit entropy-driven inhibitor interaction. Interestingly, Flap+ (I54A) exhibited significantly slower heat flow in the competitive ITC experiment with a strong binder, darunavir, and a weak binder, acetyl-pepstatin, but did not exhibit such slow heat flow in the direct inhibitor-titration experiments. NMR confirmed replacement of the weak binder by the strong binder in a competitive manner. This difference in the heat flow of the competitive binding experiment compared to the direct experiment can only be explained by assuming an inhibitor-bound intermediate pathway. A similar, but attenuated, tendency for slow heat flow was also detected in the competitive experiment with WT. Overall, our data suggests that an inhibitor-bound intermediate affects the entropy-enthalpy compensation of inhibitor-PR interaction.