In a large placebo-controlled efficacy trial of the rhesus tetravalent (RRV-TV) and serotype 1 monovalent (RRV-S1) rotavirus vaccines in multiple sites throughout the United States, protection against rotavirus disease over a 2-year period was found to be 57 and 40%, respectively (Bernstein et al., J. Am. Med. Assoc., 1995, 273, 1191–1196). Sera collected from a subset of subjects during this trial were used to determine possible correlations between rotavirus antibody responses after vaccination and protection. Between 82% (RRV-S1) and 92% (RRV-TV) of the vaccinees seroconverted by at least one of the six antibody assays performed (i.e. rotavirus IgA and neutralizing antibody to RRV and serotype 1–4 human rotaviruses). Rises in neutralizing antibody were due primarily to RRV. The seroconversion rate was only 18–22% to each of the four human rotavirus serotypes following RRV-TV vaccination and was only 43% to serotype 1 human rotavirus after RRV-S1 administration. Furthermore, no correlate of immunity against rotavirus infection or disease was identifiable based on seroconversion to any of the antibodies measured. Likewise, no consistent relationship was found between the titers of any of these six antibodies following vaccination and protection against rotavirus, thus suggesting that serum antibody titers will not be useful markers of protection with these reassortant RRV vaccines. In addition, vaccinated subjects did not develop higher titers of neutralizing antibody to human rotaviruses following a subsequent natural rotavirus illness, a further indication that only weak immune responses to human rotaviruses were stimulated by vaccination with the RRV reassortants. Finally, vaccinees in this study were found to have reduced numbers of total as well as symptomatic rotavirus infections, possibly because the vaccines shifted what would have been symptomatic infections to asymptomatic infections and what would have been asymptomatic infections to complete protection.