Methods for detecting norovirus (NoV) in food are crucial for investigation and prevention of outbreaks caused by NoV-contaminated food. However, current NoV detection methods have not been well examined or optimized. In this study, the effectiveness of various methods for eluting NoV from various fruit, concentrating the virus using polyethylene glycol (PEG), and extracting the viral RNA for subsequent assay by RT-PCR was optimized. First, six different buffers previously described for eluting NoV from fruit surfaces were evaluated. A known amount of NoV was spiked onto the surface of grapes, strawberries, and raspberries, and the virus was recovered with distilled water, 0.05 M glycine-0.14 M NaCl (pH 7.5), 2.9% tryptose phosphate broth-6% glycine, 100 mM Tris-HCl (pH 9.5), 50 mM glycine-50 mM MgCl(2) (pH 9.5), or 3% beef extract. Quantitation of the recovered virus using RT-PCR revealed that the most effective elution buffer was 3% beef extract. Secondly, to optimize a method for concentrating the recovered NoV, the key parameters of PEG precipitation, a typical method for concentrating enteric virus, were investigated. The influence of PEG molecular weight and the duration and temperature of the precipitation procedure were examined. NoV was concentrated most efficiently by precipitation when PEG (10,000) was used for 4h at room temperature. Finally, five different methods for nucleic acid extraction were evaluated. Among RNA extraction methods examined, QIAamp Viral RNA Mini kit showed the best recovery efficiency. Using the optimized method, approximately 6-80% of the seeded NoV was recovered from the various fruit.