Objectives <p>Autoimmune encephalitis (AE) is a major cause of acute encephalopathy in children, and relapse is strongly linked to long-term neurodevelopmental sequelae. However, studies on relapse risk prediction in the Chinese pediatric population remain limited, and validated tools for individualized risk stratification are lacking. This study aimed to develop and validate a nomogram for predicting relapse risk in children with AE.</p> Methods <p>This retrospective cohort study included 127 children aged 1–16 years diagnosed with AE at a tertiary pediatric center in China between 2017 and 2023. Independent predictors of relapse were identified using univariable and multivariable logistic regression. A nomogram was constructed and comprehensively validated for discrimination, calibration, and stability.</p> Results <p>The overall relapse rate was 26.8% (34/127). Independent risk factors for relapse included age ≥ 8 years at onset (OR 3.19, 95% CI 1.37–7.42, P 0.007), limb weakness (OR 2.36, 95% CI 1.03–5.42, P 0.041), delayed immunotherapy (OR 2.77, 95% CI 1.23–6.24, P 0.013), and positive N-methyl-D-aspartate receptor (NMDAR) antibody (OR 2.91, 95% CI 1.29–6.56, P 0.010). Rituximab treatment (OR 0.43, 95%CI 0.19–0.95, P 0.036) and IVIG treatment ≥ 6 months (OR 0.35, 95%CI 0.13–0.92, P 0.033) were confirmed as independent protective factors. The nomogram showed excellent discriminative ability (C-statistic 0.936, R² 0.683), good calibration (Brier score 0.081, calibration intercept &lt; 0.001, slope 1.000), and robust stability on internal validation with 1000 bootstrap resamples.</p> Conclusion <p>This nomogram demonstrates favorable accuracy and clinical utility for individualized relapse risk prediction in Chinese children with AE.</p>