Human noroviruses (HuNoVs) are a major cause of foodborne gastroenteritis worldwide. Because they do not grow in cell culture and there is no animal model for HuNoVs, pathogenesis studies have been hampered. Thus, little is known about their replication strategies or induction of neutralizing antibodies.The limited information on their pathogenesis is from human volunteer studies of HuNoV infections in which villus atrophy in duodenal biopsies and presence of malabsorptive diarrhea were described (1,7,8). No information is available on lesions in other portions of the intestines of these volunteers (9). Intestinal transplant pediatric patients that were diagnosed with HuNoV infection developed secretory or osmotic diarrhea (19,20,31). These patients had prolonged diarrhea (17 to 326 days) due to immunosuppressive therapy. The detection of HuNoV RNA and the clinical symptoms remitted after reduction of the immunosuppressive therapy. Usually in exposed individuals, histologic lesions correlate with diarrhea, but in one report, lesions in volunteers who did not show clinical symptoms were described (42). There are also numerous reports of asymptomatic individuals who were infected with HuNoVs and shed virus in the feces (11, 28).Most past attempts to study these viruses in an animal model may have failed because (i) the human strains that were used were not closely related to the host animal NoV strains, (ii) sensitive detection techniques were lacking, and finally (iii) the role of histo-blood group antigen (HBGA) phenotypes in differential susceptibility of the host was unrecognized. Our goal was to adapt a HuNoV strain to replicate in the gnotobiotic (Gn) pig to develop an animal model for the study of HuNoV pathogenesis. Gnotobiotic pigs are good models for human enteric diseases (40) because pigs resemble humans in their gastrointestinal anatomy, physiology, and immune responses. The Gn pigs are immunocompetent at birth, but they lack maternal antibodies and previous or ongoing exposure to microbial agents, including caliciviruses.Recently, viral RNA genetically similar to that of human NoV GII (65 to 71% amino acid sequence identity in the capsid gene) was detected in pigs in Japan (46, 47) and Europe (22,48). In U.S. swine, our laboratory detected both viral RNA and virus particles similar to GII HuNoV (70% sequence identity in the capsid region) which were infectious for Gn pigs (50). Our approach to infect Gn pigs with a HuNoV was to use a GII strain that is closely related genetically to the identified GII porcine NoVs and that has a broad HBGA binding pattern because little information or reagents are available for pig HBGA. Additionally, we used sensitive assays and reagents including reverse transcription (RT)-PCR to detect fecal shedding, virus-like particles (VLPs) for serological assays, and antisera to these VLPs for antigen enzyme-linked immunosor-