Noroviruses are a major cause of acute gastroenteritis worldwide (1). Prompt and accurate laboratory confirmation of norovirus infections is important to identify the source and interrupt virus transmission. Norovirus antigen detection by enzyme immunoassays (EIAs) is commonly used in settings where molecular tests are not available. The performance of antigen-based EIAs is complicated, however, by the genetic diversity and evolution of noroviruses. Phylogenetically classified into at least 6 genogroups and subdivided into nearly 40 genotypes (2), noroviruses are genetically very diverse, and each genotype may represent a separate serotype (3). Furthermore, due to the low-fidelity RNAdependent RNA polymerase, novel norovirus variants emerge periodically. For example, over the last 2 decades, new genogroup II genotype 4 (GII.4) variants have emerged every 2 to 4 years (4, 5). In the winter of 2014 to 2015, a novel norovirus GII.17 Kawasaki 2014 variant replaced the predominant GII.4 Sydney 2012 variant in several Asian countries (6-9). Here, we report on the reduced diagnostic performance of two commonly used commercial norovirus EIAs for this emergent GII.17 variant.A total of 90 stool samples that tested positive for norovirus by real-time reverse transcription-PCR (RT-PCR) were randomly selected from a pool of 141 samples for this study. They consisted of 25 GII.4 Sydney 2012-positive specimens and 65 GII.17 Kawasaki 2014-positive specimens. The samples were collected on presentation from patients that were hospitalized for norovirus gastroenteritis between December 2014 and March 2015, during which time the GII.17 Kawasaki 2014 variant outnumbered the GII.4 Sydney 2012 variant in Hong Kong (7). The median age of the patients was 19 years (interquartile range, 2 to 70 years). The female-to-male ratio was 1.4. Two norovirus EIAs were evaluated, Ridascreen norovirus 3rd generation antigen EIA (R-Biopharm) and IDEIA norovirus EIA (Oxoid), referred to as Ridascreen and IDEIA, respectively. The two kits were purchased in 2015 and were used before the indicated expiration dates. Testing was performed as per the manufacturers' instructions by one experienced researcher who was blinded to the norovirus genotype information of the samples. Samples with readings in the equivocal range were counted as positive. All samples were kept frozen at Ϫ70°C before testing. Fecal norovirus RNA load was measured using a broadly reactive real-time RT-PCR assay and was expressed as a cycle threshold (C T ) value that was inversely proportional to the RNA copy number (10). The 95% confidence interval (95% CI) of clinical sensitivity was calculated using sampsize (available at http: //sampsize.sourceforge.net). Spearman's rank test was used to compare the correlations between C T values and EIA optical density readings for each EIA norovirus genotype pair. Clinical and virologic factors (age, sex, norovirus genotype, and fecal norovirus load) that might be associated with the diagnostic performance of EIAs were analyzed using multivari...