A randomised, controlled trial of bovine rotavirus vaccine was undertaken in Gambian infants. Three doses were administered, from the age of ten weeks, concurrently with oral or killed polio vaccine. Prevaccination rotavirus neutralising antibody levels were high. 84/185 infants (45%) showed an increase in neutralising antibody titre after receiving rotavirus vaccine, compared with 20/91 (22%) unvaccinated infants. Clinical rotavirus infection was detected in 24/78 (31%) children in the rotavirus/oral polio group, 34/83 (41%) children in the placebo/oral polio group, and 23/92 (25%) children in the rotavirus/killed polio group, giving an overall vaccine efficacy of 33% (95% CI 4-53%). RIT 4237 did not appear to reduce the severity of clinical infections. Most cases (92%) were caused by rotaviruses with short RNA electropherotypes. Serological responses to rotavirus vaccination appeared unaffected by the concurrent administration of oral polio vaccine. Lower types 1 and 3 polio antibody levels were found in children who received oral polio and rotavirus vaccines but the differences were not statistically significant.