Evidence indicates that fractional doses of yellow fever vaccine are safe and sufficiently immunogenic for use during yellow fever outbreaks. However, there are no data on the generalisability of this observation to populations living with HIV. Therefore, we aimed to evaluate the immunogenicity of fractional and standard doses of yellow fever vaccine in HIV-positive adults. We conducted a randomised, double-blind, non-inferiority substudy in Kilifi, coastal Kenya to compare the immunogenicity and safety of a fractional dose (one-fifth of the standard dose) versus the standard dose of 17D-213 yellow fever vaccine among HIV-positive volunteers. HIV-positive participants aged 18-59 years, with baseline CD4+ T-cell count of at least 200 cells per mL, and who were not pregnant, had no previous history of yellow fever vaccination or infection, and had no contraindication for yellow fever vaccination were recruited from the community. Participants were randomly assigned 1:1 in blocks (variable block sizes) to either a fractional dose or a standard dose of the 17D-213 yellow fever vaccine. Vaccines were administered subcutaneously by an unblinded nurse and pharmacist; all other study personnel were blinded to the vaccine allocation. The primary outcome of the study was the proportion of participants who seroconverted by the plaque reduction neutralisation test (PRNT50) 28 days after vaccination for the fractional dose versus the standard dose in the per-protocol population. Secondary outcomes were assessment of adverse events and immunogenicity during the 1-year follow-up period. Participants were considered to have seroconverted if the post-vaccination antibody titre was at least 4 times greater than the pre-vaccination titre. We set a non-inferiority margin of not less than a 17% decrease in seroconversion in the fractional dose compared with the standard dose. This study is registered with ClinicalTrials.gov, NCT02991495. Between Jan 29, 2019, and May 17, 2019, 303 participants were screened, and 250 participants were included and vaccinated; 126 participants were assigned to the fractional dose and 124 to the standard dose. 28 days after vaccination, 112 (96%, 95% CI 90-99) of 117 participants in the fractional dose group and 115 (98%, 94-100) of 117 in the standard dose group seroconverted by PRNT50. The difference in seroconversion between the fractional dose and the standard dose was -3% (95% CI -7 to 2). Fractional dosing therefore met the non-inferiority criterion, and non-inferiority was maintained for 1 year. The most common adverse events were headache (n=31 [12%]), fatigue (n=23 [9%]), myalgia (n=23 [9%]), and cough (n=14 [6%]). Reported adverse events were either mild (182 [97%] of 187 adverse events) or moderate (5 [3%]) and were self-limiting. Fractional doses of the 17D-213 yellow fever vaccine were sufficiently immunogenic and safe demonstrating non-inferiority to the standard vaccine dose in HIV-infected individuals with CD4+ T cell counts of at least 200 cells per mL. These results provide confidence that fractional dose recommendations are applicable to populations with high HIV prevalence. Wellcome Trust, Médecins Sans Frontières Foundation, and the UK Department for International Development.