Transition state and high energy intermediate mimetics have the potential to be very potent enzyme inhibitors. In this study, a model of peptide hydrolysis in the active site of insulin-regulated aminopeptidase (IRAP) was developed using density functional theory calculations and the cluster approach. The 3D structure models of the reaction coordinates were used for virtual screening to obtain new chemical starting points for IRAP inhibitors. This mechanism-based virtual screening process managed to identify several known peptidase inhibitors from a library of over 5 million compounds, and biological testing identified one compound not previously reported as an IRAP inhibitor. This novel methodology for virtual screening is a promising approach to identify new inhibitors mimicking key transition states or intermediates of an enzymatic reaction.