Outbreaks of hepatitis A virus (HAV) infections are most frequently associated with the consumption of contaminated oysters. A rapid and selective concentration method is necessary for the recovery of HAV from contaminated oysters prior to detection using PCR. In this study, ricin extracted from castor beans (Ricinus communis) was tested as an alternative to antibody used in immunomagnetic separation while concentrating HAV prior to its detection using reverse transcription PCR. Initially, the extracted proteins from castor beans were fractionated into 13 fractions by gel filtration chromatography. Pretreatment of different protein fractions showed a variation in binding of HAV viral protein (VP) 1 to oyster digestive tissue in the range of 25.9 to 63.9%. The protein fraction, which caused the highest reduction in binding of VP1 to the tissue, was identified as ricin A by quadrupole time-of-flight mass spectrometry. Ricin A could significantly inhibit binding of VP1 to the tissue with a 50% inhibitory concentration of 4.5 μg/ml and a maximal inhibitory concentration of 105.2%. The result showed that the rate of inhibition of HAV binding to tissue was higher compared to the rate of ricin itself binding to HAV (slope: 0.0029 versus 0.00059). However, ricin A concentration showed a higher correlation to the relative binding of ricin itself to HAV than the inhibition of binding of HAV to the tissue (coefficient of determination, R(2): 0.9739 versus 0.6804). In conclusion, ricin A-linked magnetic bead separation combined with reverse transcription PCR can successfully detect HAV in artificially seeded oyster digestive tissue up to a 10(-4) dilution of the virus stock (titer: 10(4) 50% tissue culture infective dose per ml).