A randomized controlled trial was performed in infants undergoing routine immunization in North Hertfordshire. Ninety-six children received a single dose of inactivated polio vaccine, followed by two doses of live attenuated oral polio vaccine and 97 children received three doses of live attenuated oral polio vaccine at 2, 3 and 4 months of age. Blood samples were taken by study nurses 6 weeks after vaccination and stool samples were collected by parents weekly for 4 weeks after each dose of vaccine. Follow-up was completed for 92 of 96 (96%) children in the combined schedule group and 92 of 97 (95%) in the control group. After vaccination the proportions of children with detectable antibody to poliovirus serotypes 1, 2 and 3 were high and similar between groups and geometric mean titers (95% confidence interval) to poliovirus types 1, 2 and 3 were 264 (200 to 347), 375 (311 to 450) and 189 (144 to 250) in the combined schedule group and 369 (290 to 469), 401 (321 to 498) and 206 (145 to 293) in the live vaccine group, respectively. The only significant difference between groups in rates of viral excretion was observed after the second dose of live attenuated oral polio vaccine, when excretion of type 3 poliovirus was reduced in those children who had received piror inactivated polio vaccine (P= 0.05). This study suggests that, compared with the current schedule, a combined schedule of inactivated and live polivaccines is likely to produce equivalent individual protection against poliomyelitis and is unlikely to substantially alter circulation of poliovirus in the community. Becuase the risk of vaccine-associated polio-myelitis is greatest after the first dose of poliovaccine, this schedule could also reduce the risk to vaccine recipients.