Our studies of rotavirus challenge in adult volunteers enabled us to evaluate the relationship of pre-existing antirotavirus fecal IgA antibody to infection and illness and to investigate the local response to this infection. No relationship could be found between the pre-existing levels of fecal antirotavirus IgA antibody and protection from infection or illness. A greater than six-fold increase in the level of antibody was seen in 16/19 infected volunteers with determinable increases but in 0/15 controls who received less than the minimal infectious dose of rotavirus. Antibody levels increased rapidly in infected volunteers and were consistent with an anamnestic response. Two of seven volunteers who received an infectious dose of rotavirus but were considered uninfected on the basis of other laboratory methods had greater than or equal to six-fold increases of fecal antibody and one of these experienced symptoms compatible with a rotavirus infection. This finding indicates that an increase in fecal antibody may be a reliable indicator of rotavirus infection even in the absence of detectable shedding or seroconversion.