The design and synthesis of novel, orally active, potent, and selective inhibitors of influenza neuraminidase differing structurally from existing neuraminidase inhibitors are described. X-ray crystal structures of complexes of neuraminidase with known five- and six-membered ring inhibitors revealed that potent inhibition of the enzyme is determined by the relative positions of the interacting inhibitor substituents (carboxylate, glycerol, acetamido, hydroxyl) rather than by the absolute position of the central ring. This led us to design potential neuraminidase inhibitors in which the cyclopentane ring served as a scaffold for substituents (carboxylate, guanidino, acetamido, alkyl) that would interact with the four binding pockets of the neuraminidase active site at least as effectively as those of the established six-membered ring inhibitors such as DANA (2), zanamivir (3), and oseltamivir (4). A mixture of the isomers was prepared initially. Protein crystallography of inhibitor-enzyme complexes was used to screen mixtures of isomers in order to identify the most active stereoisomer. A synthetic route to the identified candidate 50 was developed, which featured (3 + 2) cycloaddition of 2-ethylbutyronitrile oxide to methyl (1S,4R)-4[(tert-butoxycarbonyl)amino]cyclopent-2-ene-1-carboxylate (43). Structures of the synthetic compounds were verified by NMR spectroscopy using nuclear Overhauser effect methodology. Two new neuraminidase inhibitors discovered in this work, 50 and 54, have IC(50) values vs neuraminidase from influenza A and B of <1 and <10 nM, respectively. These IC(50) values are comparable or superior to those for zanamivir and oseltamivir, agents recently approved by the FDA for treatment of influenza. The synthetic route used to prepare 50 and 54 was refined so that synthesis of pure active isomer 54, which has five chiral centers, required only seven steps from readily available intermediates. Further manipulation was required to prepare deoxy derivative 50. Because the activities of the two compounds are comparable and 54 [RWJ-270201 (BCX-1812)] is the easier to synthesize, it was selected for further clinical evaluation.