BACKGROUND: Following the global eradication of wild poliovirus, countries using live attenuated oral poliovirus vaccines will transition to exclusive use of inactivated poliovirus vaccine (IPV) or fractional doses of IPV (f-IPV; a f-IPV dose is one-fifth of a normal IPV dose), but IPV supply and cost constraints will necessitate dose-sparing strategies. We compared immunisation schedules of f-IPV and IPV to inform the choice of optimal post-eradication schedule. METHODS: This randomised open-label, multicentre, phase 3, non-inferiority trial was done at two centres in Panama and one in the Dominican Republic. Eligible participants were healthy 6-week-old infants with no signs of febrile illness or known allergy to vaccine components. Infants were randomly assigned (1:1:1:1, 1:1:1:2, 2:1:1:1), using computer-generated blocks of four or five until the groups were full, to one of four groups and received: two doses of intradermal f-IPV (administered at 14 and 36 weeks; two f-IPV group); or three doses of intradermal f-IPV (administered at 10, 14, and 36 weeks; three f-IPV group); or two doses of intramuscular IPV (administered at 14 and 36 weeks; two IPV group); or three doses of intramuscular IPV (administered at 10, 14, and 36 weeks; three IPV group). The primary outcome was seroconversion rates based on neutralising antibodies for poliovirus type 1 and type 2 at baseline and at 40 weeks (4 weeks after the second or third vaccinations) in the per-protocol population to allow non-inferiority and eventually superiority comparisons between vaccines and regimens. Three co-primary outcomes concerning poliovirus types 1 and 2 were to determine if seroconversion rates at 40 weeks of age after a two-dose regimen (administered at weeks 14 and 36) of intradermally administered f-IPV were non-inferior to a corresponding two-dose regimen of intramuscular IPV; if seroconversion rates at 40 weeks of age after a two-dose IPV regimen (weeks 14 and 36) were non-inferior to those after a three-dose IPV regimen (weeks 10, 14, and 36); and if seroconversion rates after a two-dose f-IPV regimen (weeks 14 and 36) were non-inferior to those after a three-dose f-IPV regimen (weeks 10, 14, and 36). The non-inferiority boundary was set at -10% for the lower bound of the two-sided 95% CI for the seroconversion rate difference.. Safety was assessed as serious adverse events and important medical events. This study is registered on ClinicalTrials.gov, NCT03239496. FINDINGS: From Oct 23, 2017, to Nov 13, 2018, we enrolled 773 infants (372 [48%] girls) in Panama and the Dominican Republic (two f-IPV group n=217, three f-IPV group n=178, two IPV group n=178, and three IPV group n=200). 686 infants received all scheduled vaccine doses and were included in the per-protocol analysis. We observed non-inferiority for poliovirus type 1 seroconversion rate at 40 weeks for the two f-IPV dose schedule (95·9% [95% CI 92·0-98·2]) versus the two IPV dose schedule (98·7% [95·4-99·8]), and for the three f-IPV dose schedule (98·8% [95·6-99·8]) versus the three IPV dose schedule (100% [97·9-100]). Similarly, poliovirus type 2 seroconversion rate at 40 weeks for the two f-IPV dose schedule (97·9% [94·8-99·4]) versus the two IPV dose schedule (99·4% [96·4-100]), and for the three f-IPV dose schedule (100% [97·7-100]) versus the three IPV dose schedule (100% [97·9-100]) were non-inferior. Seroconversion rate for the two f-IPV regimen was statistically superior 4 weeks after the last vaccine dose in the 14 and 36 week schedule (95·9% [92·0-98·2]) compared with the 10 and 14 week schedule (83·2% [76·5-88·6]; p=0·0062) for poliovirus type 1. Statistical superiority of the 14 and 36 week schedule was also found for poliovirus type 2 (14 and 36 week schedule 97·9% [94·8-99·4] vs 10 and 14 week schedule 83·9% [77·2-89·2]; p=0·0062), and poliovirus type 3 (14 and 36 week schedule 84·5% [78·7-89·3] vs 10 and 14 week schedule 73·3% [65·8-79·9]; p=0·0062). For IPV, a two dose regimen administered at 14 and 36 weeks (99·4% [96·4-100]) was superior a 10 and 14 week schedule (88·9% [83·4-93·1]; p<0·0001) for poliovirus type 2, but not for type 1 (14 and 36 week schedule 98·7% [95·4-99·8] vs 10 and 14 week schedule 95·6% [91·4-98·1]), or type 3 (14 and 36 week schedule 97·4% [93·5-99·3] vs 10 and 14 week schedule 93·9% [89·3-96·9]). There were no related serious adverse events or important medical events reported in any group showing safety was unaffected by administration route or schedule. INTERPRETATION: Our observations suggest that adequate immunity against poliovirus type 1 and type 2 is provided by two doses of either IPV or f-IPV at 14 and 36 weeks of age, and broad immunity is provided with three doses of f-IPV, enabling substantial savings in cost and supply. These novel clinical data will inform global polio immunisation policy for the post-eradication era. FUNDING: Bill & Melinda Gates Foundation.