Factors such as low recovery rate and food contaminants may be responsible for the di$culty of detecting Norovirus (NV) by PCR in foodborne outbreaks. To detect NV more e$ciently, we introduced a bacterial treatment, in which concentrated samples were incubated overnight with Klebsiella oxytoca at 35ῒ before RNA extraction using the standard protocol. Recovery rates of NVs (G I/8 or G II/13) added to food suspensions in the modified method were compared with those in the standard method by quantification of NV RNAs using real-time PCR. Recovery rates in the modified method were 8.6ΐ for G I/8 and 11.6ΐ for G II/13 in 18 oyster samples and 13.9ΐ for G I/8 and 19.6ΐ for G II/13 in 15 other food samples, while those in the standard method were 0.3ΐ for G I/8 and 0.5ΐ for G II/13 in the oyster samples and 1.9ΐ for G I/8 and 7.9ΐ for G II/13 in the other food samples. These results indicate that the bacterial treatment increase the recovery of NV from foods such as oysters, suggesting that the modified method will be useful for NV detection in food samples.