The neuraminidase of influenza virus is a surface glycoprotein that catalyzes the hydrolysis of glycosidic linkages between terminal sialic acids and adjacent sugar moieties. Neuraminidase function is critical for the spread of virus to new cells, and if the enzyme activity is inhibited, then virus infection is abrogated. The neuraminidase active site is conserved in all influenza type-A and type-B isolates, which makes it an excellent target for drug design. To determine the potential for resistance to develop against neuraminidase inhibitors, we have constructed mutations in seven of the conserved active-site residues of a type B (B/Lee/40) neuraminidase and analyzed the effect of the altered side chains on enzyme activity. There is a reduction in k(cat) in all our mutants. A transition-state analogue inhibitor shows variation in Ki with the mutant neuraminidases, allowing interpretation of the effects of mutation in terms of transition-state binding and product release. The results show that Tyr409 is the most critical residue for enzyme activity, but that Asp149, Arg223, Glu275 and Arg374 also play important roles in enzyme catalysis. Based on the pH profile of neuraminidase activity of the D149E mutant protein, we conclude that Asp149 is not a proton donor, but is involved in stabilizing the transition state. If designed inhibitors are targeted to these residues where mutations are highly deleterious, particularly Tyr409, Glu275 and Asp149, the virus is unlikely to generate resistance to the drug.