To explore the clinical characteristics of recurrence in pediatric epilepsy after treatment and analyze the electroencephalogram (EEG) features during the interictal period in epileptic children. Methods: A retrospective analysis was conducted on 168 pediatric patients with epilepsy admitted to our hospital. The patients were divided into a recurrence group (n=64) and a non-recurrence group (n=104) based on follow-up results. A multivariate Logistic regression model was used to analyze the independent influencing factors of prognosis recurrence in the patients, as well as the correlation between EEG parameters and clinical characteristics of epilepsy recurrence. A Bayesian network was employed to construct a prediction model for epilepsy recurrence in the patients. The Kaplan-Meier survival curve model was utilized to draw the cumulative recurrence curve during the follow-up period and analyze the recurrence situation of the patients during the follow-up. Results: Male gender, neurological dysfunction, sleep disorders, time to effective control exceeding one year, seizure frequency exceeding five times per year, MRI showing intracranial lesions, negative FAT1 expression, and anxiety were identified as independent risk factors for recurrence in the patients (P<0.05). Independent protective factors for recurrence included no seizure for at least five years before medication reduction, medication reduction duration exceeding six months, age of onset between 3 and 14 years, and α, θ, and δ bands after medication withdrawal (P<0.05). The α, θ, and δ waves were lower in the recurrence group compared to the non-recurrence group after medication withdrawal (P<0.05). There was a significant positive correlation between medication reduction duration and the frequencies of α, θ, and δ bands (P<0.05), while sleep disorders, time to effective control, seizure frequency, and neurological dysfunction were significantly negatively correlated with the frequencies of α, θ, and δ bands (P<0.05). Most recurrences occurred within 24 months after treatment. Conclusion: Male gender, neurological dysfunction, sleep disorders, time to effective control exceeding one year, seizure frequency exceeding five times per year, MRI showing intracranial lesions, negative FAT1 expression, and anxiety are independent risk factors for recurrence in pediatric patients with epilepsy. Independent protective factors for recurrence include no seizure for at least five years before medication reduction, medication reduction duration exceeding six months, age of onset between 3 and 14 years, and α, θ, and δ bands after medication withdrawal.