HIV-1 protease is the most effective target for drugs to treat AIDS, however, the long-term therapeutic efficiency is restricted by the rapid development of drug resistant variants. To better understand the molecular basis of drug resistance, crystallographic and kinetic studies were applied to wild-type HIV-1 protease (PR) and drug-resistant mutants, PRV82A, and PRI84V, in complex with substrate analogues, the current drug saquinavir and the new inhibitor UIC-94017 (TMC-114). UIC-94017 was also studied with mutants PRD30N and PRI50V. The drug-resistant mutations V82A, I84V, D30N and I50V participate in substrate binding. Eighteen crystal structures were refined at resolutions of 0.97-1.60A. The high accuracy of the atomic resolution crystal structures helps understand the reaction mechanism of HIV-1 PR. Different binding modes are observed for different types of inhibitors. The substrate analogs have more extended interactions with PR subsites up to S5-S5', while the clinical inhibitors maximize the contacts within S2-S2'. Hydrophobic interactions are the major force for saquinavir binding since it was designed with enhanced hydrophobic groups based on substrate side-chains. In contrast, the new clinical inhibitor UIC-94017 was designed to mimic the hydrogen bonds between substrates and PR. UIC-94017 forms polar interactions with the PR main-chain atoms of Asp29/30, which have been proposed to be critical for its potency against resistant HIV. The mutants showed different structural and kinetic effects, depending on the inhibitor and location of the mutations. The observed structural changes were consistent with the relative inhibition data. Both PRI84V and PRI50V lost favorable hydrophobic interactions with inhibitor compared with PR. Similarly, in PRD30N the UIC-94017 had a water-mediated interaction with the side-chain of Asn30 rather than the direct interaction observed in PR. However, PRV82A compensated for the mutation by shifts of the backbone of Ala82. Furthermore, the complexes of PRV82A showed smaller shifts relative to PR, but more movement of the peptide analog, compared to complexes with clinical inhibitors. The structures suggest that substrate analogs have more flexibility than the drugs to accommodate the structural changes caused by mutation, which may explain how HIV can develop drug resistance while retaining the ability of PR to hydrolyze natural substrates.