Data on long-term neurodevelopmental outcomes of normocephalic children (born with normal head circumference) exposed to Zika virus in utero are scarce. We aimed to compare neurodevelopmental outcomes in normocephalic children up to age 48 months with and without Zika virus exposure in utero. In this prospective cohort study, we included infants from two cohorts of normocephalic children born in León and Managua, Nicaragua during the 2016 Zika epidemic. In León, all women pregnant during the two enrolment periods were eligible. In Managua, mother-child pairs were included from three districts in the municipality of Managua: all women who became pregnant before June 15, 2016, and had a due date of Sept 15, 2016 or later were eligible. Infants were serologically classified as Zika virus-exposed or Zika virus-unexposed in utero and were followed up prospectively until age 48 months. At 36 months and 48 months of age, the Mullen Scales of Early Learning (MSEL) assessment was administered. Primary outcomes were MSEL early learning composite (ELC) scores at 30-48 months in León and 36-48 months in Managua. We used an inverse probability weighting generalised estimating equations model to assess the effect of Zika virus exposure on individual MSEL cognitive domain scores and ELC scores, adjusted for maternal education and age, poverty status, and infant sex. The initial enrolment period for the León cohort was between Jan 31 and April 5, 2017 and the second was between Aug 30, 2017, and Feb 22, 2018. The enrolment period for the Managua cohort was between Oct 24, 2019, and May 5, 2020. 478 mothers (482 infants) from the León cohort and 615 mothers (609 infants) from the Managua cohort were enrolled, of whom 622 children (303 from the León cohort; 319 from the Managua cohort) were included in the final analysis; four children had microcephaly at birth and thus were excluded from analyses, two from each cohort. 33 (11%) of 303 children enrolled in León and 219 (69%) of 319 children enrolled in Managua were exposed to Zika virus in utero. In both cohorts, no significant differences were identified in adjusted mean ELC scores between Zika virus-exposed and unexposed infants at 36 months (between-group difference 1·2 points [95% CI -4·2 to 6·5] in the León cohort; 2·8 [-2·4 to 8·1] in the Managua cohort) or at 48 months (-0·9 [-10·8 to 8·8] in the León cohort; 0·1 [-5·1 to 5·2] in the Managua cohort). No differences in ELC scores between Zika virus-exposed and unexposed infants exceeded 6 points at any time between 30 months and 48 months in León or between 36 months and 48 months in Managua, which was considered clinically significant in other settings. We found no significant differences in neurodevelopmental scores between normocephalic children with in-utero Zika virus exposure and Zika virus-unexposed children at age 36 months or 48 months. These findings are promising, supporting typical neurodevelopment in Zika virus-exposed normocephalic children, although additional follow-up and research is warranted. National Institute of Child Health and Development, National Institute of Allergy and Infectious Diseases, and Fogarty International Center. For the Spanish translation of the abstract see Supplementary Materials section.