BACKGROUND: A rapid increase in circulating vaccine-derived poliovirus type 2 outbreaks, and the need to reserve inactivated poliovirus vaccine (IPV) for routine immunisation, has increased the value of fractional dose IPV (fIPV) as a measure to prevent acute flaccid paralysis. However, the intradermal route of administration has been viewed as prohibitive to outbreak response campaigns. We aimed to establish the immunogenicity and safety of administering intradermal fIPV with a disposable syringe jet injector (DSJI) or an intradermal adaptor (IDA) compared with standard administration with a BCG needle and syringe (N&S). METHODS: This pragmatic, non-inferiority trial was undertaken in a campaign setting in communities in The Gambia. Children aged 4-59 months without contraindication to vaccination were eligible. Children were not individually randomly assigned; instead, the vaccination teams were randomly assigned (1:1:1) to one of three administration methods. Parents and the field team were not masked, but laboratory personnel were masked. Baseline demographic and anthropometric data were collected from the participants. Public health officers experienced at intradermal immunisation, and nurses without experience, had 2 h of training on each of the administration methods before the campaign. Participants were vaccinated using the administration method in use by the vaccination team in their community. Poliovirus serum neutralising antibodies (SNA) were measured in children aged 24-59 months before and 4 weeks after vaccination. Adverse events and data on injection quality were collected from all participants. The primary outcome was the type 2 immune response rate (seroconversion in seronegative [SNA titre <8] children plus a 4-fold titre rise in seropositive children). Adjusted differences in the immune response between the DSJI or IDA group versus the N&S group were calculated with 97·5% CIs. A margin of -10% was used to define the non-inferiority of DSJI or IDA compared to N&S. Immunogenicity analysis was done per protocol. The trial is registered with ClinicalTrials.govNCT02967783 and has been completed. FINDINGS: Between Oct 28 and Dec 29, 2016, 3189 children aged 4-59 months were recruited, of whom 3170 were eligible. Over 3 days, 2720 children were vaccinated (N&S, 917; IDA, 874; and DSJI, 929). Among 992 children aged 25-59 months with a baseline SNA available, 90·1% (95% CI 86·1-92·9; 281/312) of those vaccinated using the DSJI had an immune response to type 2 compared with 93·8% (90·6-95·8; 331/353) of those vaccinated with N&S and 96·6% (94·0-98·0; 316/327) of those vaccinated with IDA. All (53/53) type 2 seronegative children seroconverted. For polio type 2, non-inferiority was shown for both the IDA (adjusted difference 0·7% [97·5% CI -3·3 to 4·7], unadjusted difference 2·9% [-0·9 to 6·8]) and DSJI (adjusted difference -3·3% [-8·3 to 1·5], unadjusted difference -3·7% [-8·7 to 1·1]) compared with N&S. Non-inferiority was shown for type 1 and 3 for the IDA and DSJI. Neither injection quality nor the training and experience of the vaccinators had an effect on immune response. No safety concerns were reported. INTERPRETATION: In a campaign, intradermal fIPV is safe and generates consistent immune responses that are not dependent on vaccinator experience or injection quality when administered using an N&S, DSJI, or IDA. Countries facing vaccine-derived poliovirus type 2 outbreaks should consider fIPV campaigns to boost population immunity and prevent cases of acute flaccid paralysis. FUNDING: World Health Organization and the Medical Research Council.