In phase I and II double-blind studies in Venezuela, the rhesus rotavirus (serotype 3) candidate vaccine was fed to over 140 infants (1-10 months old) and a similar number received placebo. Significant reactions were not observed. Overall, 69.3% of the vaccinees developed a seroresponse to rhesus rotavirus (RRV). During a similar study in newborn infants reactions were not observed after feeding a full dose (104 PFU) of vaccine. Although only 2 of 20 vaccinees developed a sero-response (as detected by neutralization assay) shedding of virus was observed in 19. Phase I studies have also been performed with two RRV reassortants representing serotypes 1 (D x RRV) and 2 (DS1 x RRV). In addition a group of 23 children received a bivalent preparation (RRV and D x RRV). Minor reactions (low grade fever 2, 3 or 4 days after vaccination) were seen in 10-25% of the children in these three groups as well as in a comparison group vaccinated with RRV alone. Shedding of virus was detected in over 67% of the vaccinees. Most of the children who received RRV and D x RRV shed both viruses in their stool. During a 1 year period of longitudinal follow-up of 247 children in the first RRV study, 5 episodes of rotavirus diarrhea were observed among the vaccinees versus 16 in the placebo group for an overall vaccine efficacy of 68%. Vaccine efficacy was 93% in infants who were 1-5 months old at the time of vaccination. The six most severe episodes occurred in the placebo group for a vaccine efficacy of 100% against severe diarrhea. The rotavirus strains recovered from ill infants or children were predominantly serotype 3 and thus the protection provided by the RRV vaccine was homotypic.