Background: The longitudinal relationship between HLA class I and II eplet mismatches, de novo donor-specific antibodies (dnDSA) development, and acute rejection after transplantation in childhood is unknown. Methods: Eplet mismatches at HLA class I and II loci were calculated retrospectively for each donor/recipient pair transplanted between 2005 and 2015 at a single Australian center. Logistic regression analyses were conducted to determine the association between the number of eplet mismatches, dnDSA, and acute rejection. Results: The cohort comprised 59 children (aged 0–18 years) who received their first kidney allograft and were followed for median (interquartile range) 4.5 (± 2.6) years. Overall, 32% (19/59) developed dnDSA (class I 3% (2/59), class II 14% (8/59), 15% class I and II (9/59)), and 24% (14/59) developed biopsy-proven acute rejection. Every unit increase in class I and II eplet mismatches corresponded to an increase in risk of class I (odds ratio (OR) 1.22, 95% CI 1.07–1.39, p < 0.01) and class II (OR 1.06, 95% CI 1.01–1.11, p = 0.02) dnDSA development. Compared with recipients without dnDSA, class I and II dnDSA were associated with direction of effect towards increased risk of acute cellular rejection (class I: OR 5.87, 95% CI 0.99–34.94, p = 0.05; class II: OR 12.00, 95% CI 1.25–115.36, p = 0.03) and acute antibody-mediated rejection (class I: OR 25.67, 95% CI 3.54–186.10, p < 0.01; class II: OR 9.71, 95% CI 1.64–57.72, p = 0.01). Conclusions: Increasing numbers of HLA class I or II eplet mismatches were associated with the development of dnDSA. Children who developed dnDSA were also more likely to develop acute rejection compared with children without dnDSA.