Intradermal administration of fractional inactivated poliovirus vaccine (fIPV) is a dose-sparing alternative to the intramuscular full dose. We aimed to compare the immunogenicity of two fIPV doses versus one IPV dose for routine immunisation, and also assessed the immunogenicity of an fIPV booster dose for an outbreak response. We did an open-label, randomised, controlled, inequality, non-inferiority trial in two clinics in Dhaka, Bangladesh. Healthy infants were randomly assigned at 6 weeks to one of four groups: group A received IPV at age 14 weeks and IPV booster at age 22 weeks; group B received IPV at age 14 weeks and fIPV booster at age 22 weeks; group C received IPV at age 6 weeks and fIPV booster at age 22 weeks; and group D received fIPV at 6 weeks and 14 weeks and fIPV booster at age 22 weeks. IPV was administered by needle-syringe as an intramuscular full dose (0·5 mL), and fIPV was administered intradermally (0·1 mL of the IPV formulation was administered using the 0·1 mL HelmJect auto-disable syringe with a Helms intradermal adapter). Both IPV and fIPV were administered on the outer, upper right thigh of infants. The primary outcome was vaccine response to poliovirus types 1, 2, and 3 at age 22 weeks (routine immunisation) and age 26 weeks (outbreak response). Vaccine response was defined as seroconversion from seronegative (<1:8) at baseline to seropositive (≥1:8) or four-fold increase in reciprocal antibody titres adjusted for maternal antibody decay and was assessed in the modified intention-to-treat population (infants who received polio vaccines per group assignment and polio antibody titre results to serotypes 1, 2, and 3 at 6, 22, 23, and 26 weeks of age). The non-inferiority margin was 12·5%. This trial is registered with ClinicalTrials.gov, number NCT02847026. Between Sept 1, 2016 and May 2, 2017, 1076 participants were randomly assigned and included in the modified intention-to-treat analysis: 271 in Group A, 267 in group B, 268 in group C, and 270 in group D. Vaccine response at 22 weeks to two doses of fIPV (group D) was significantly higher (p<0·0001) than to one dose of IPV (groups A and B) for all three poliovirus serotypes: the type 1 response comprised 212 (79% 95% CI 73–83) versus 305 (57% 53–61) participants, the type 2 response comprised 173 (64% 58–70) versus 249 (46% 42–51) participants, and the type 3 response comprised 196 (73% 67–78) versus 196 (36% 33–41) participants. At 26 weeks, the fIPV booster was non-inferior to IPV (group B vs group A) for serotype 1 (−1·12% 90% CI −2·18 to −0·06), serotype 2 (0·40%, –2·22 to 1·42), and serotype 3 (1·51% –3·23 to −0·21). Of 129 adverse events, 21 were classified as serious including one death; none were attributed to IPV or fIPV. fIPV appears to be an effective dose-sparing strategy for routine immunisation and outbreak responses. US Centers for Disease Control and Prevention.