We conducted a clinical trial of rhesus rotavirus vaccine MMU-18006 (RRV, serotype 3) to assess the immunogenicity, transmissibility and booster effect of this vaccine in a welfare nursery in Sapporo, from September 1986 to October 1988. After the trial, in March 1989, an outbreak of gastroenteritis due to a wild strain of serotype 1 rotavirus (RV-1) occurred in the study population. Infants were divided into three groups based on vaccination history: five booster vaccinees, 18 one-dose vaccinees and 18 control infants who did not receive vaccine. There was a significant relationship between asymptomatic infection and higher levels of preoutbreak antibody titres against KU (serotype 1) but not RRV. Significant protection from rotavirus illness was observed both in the booster vaccine group and in the one-dose vaccine group but not in the control group. Rotavirus-specific serum IgA immune response was considered to be one of the indicators of recent rotavirus infection, and did not correlate with resistance to rotavirus illness. Our results revealed that protection from rotavirus illness was serotype-specific and that previous rotavirus infection, including vaccination, was important to induce the heterotypic immune response, and that ageing or booster inoculation of RRV might play a role in the protection against serotype 1 rotavirus infection. From our findings, a booster administration was thought to be important to induce effective heterotypic immunity and should be included in a future rotavirus vaccine trial to obtain sufficient protection against four major serotypes of rotavirus.