Determination of the sensitivity of influenza viruses to neuraminidase (NA) inhibitors is presently based on assays of NA function because, unlike available cell culture methods, the results of such assays are predictive of susceptibility in vivo. At present the most widely used substrate in assays of NA function is the fluorogenic reagent 2'-O-(4-methylumbelliferyl)-N-acetylneuraminic acid (MUN). A rapid assay with improved sensitivity is required because a proportion of clinical isolates has insufficient NA to be detectable in the current fluorogenic assay, and because some mutations associated with resistance to NA inhibitors reduce the activity of the enzyme. A chemiluminescence-based assay of NA activity has been developed that uses a 1,2-dioxetane derivative of sialic acid (NA-STAR) as the substrate. When compared with the fluorogenic assay, use of the NA-STAR substrate results in a 67-fold reduction in the limit of detection of the NA assay, from 200 pM (11 fmol) NA to 3 pM (0.16 fmol) NA. A panel of isolates from phase 2 clinical studies of zanamivir, which were undetectable in the fluorogenic assay, was tested for activity using the NA-STAR substrate. Of these 12 isolates with undetectable NA activity, 10 (83%) were found to have detectable NA activity using the NA-STAR substrate. A comparison of sensitivity to zanamivir of a panel of influenza A and B viruses using the two NA assay methods has been performed. IC(50) values for zanamivir using the NA-STAR were in the range 1.0-7.5 nM and those for the fluorogenic assay in the range 1. 0-5.7 nM (n = 6). The NA-STAR assay is a highly sensitive, rapid assay of influenza virus NA activity that is applicable to monitoring the susceptibility of influenza virus clinical isolates to NA inhibitors.