Objective To investigate the clinical characteristics and prognostic factors of children with relapsed acute lymphoblastic leukemia (ALL). Methods A total of 458 newly diagnosed ALL children who treated with the Chinese Children's Leukemia Group (CCLG) protocol at hospital between February 2006 and December 2019 were selected. Results The overall relapse rate of childhod ALL in this center was 16.6% (76/458). The mortality rate among relapsed ALL children was 57.9% (44/76), and the 5-year overall survival (OS) rate for relapsed ALL children was 38.6% ± 5.9%. Grouped by time to relapse, the cohort included 26 cases of very early relapse, 30 cases of early relapse, and 20 cases of late relapse. There was a statistically significant difference in the 5-year overall survival rate (OS) among the three groups(P<0.001). When categorized by relapse site, 57 cases involved isolated bone marrow relapse, 12 cases had extramedullary relapse, and 7 cases exhibited combined medullary/extramedullary relapse. There was a statistically significant difference in the 5-year OS rate among the three groups (P<0.05). Among the 76 relapsed children, 11 discontinued treatment, while 65 received retreatment. Among them, 14 failed to achieve a second complete remission (CR2), whereas 51 attained CR2. There was a statistically significant difference in the 5-year OS rate between the two groups (P<0.001). Based on post-relapse treatment modalities,the patients were divided into allogeneic hematopoietic stem cell transplantation (Allo-HSCT) group (22 cases, 33.8%), chimeric antigen receptor T-cell immunotherapy (CART) group (8 cases, 12.3%), CAR-T combined with Allo-HSCT group (14 cases, 21.5%), and chemotherapy and/or targeted therapy group (21 cases, 32.2%). There was a statistically significant difference in the 5-year OS rate among the groups (P<0.001). Univariate prognostic analysis revealed that initial white blood cell count>100×10⁹/L, initial risk stratification, relapse time, relapse site, post-relapse risk stratification, post-relapse treatment modality, and failure to achieve CR2 were independent risk factors affecting the prognosis of relapsed ALL children (P<0.05). Multivariate analysis using the Cox regression model identified very early relapse and failure to attain CR2 after relapse as independent risk factors for poor prognosis in relapsed ALL children (P<0.05), while post-relapse treatment with CAR-T bridging to allogeneic hematopoietic stem cell transplantation (Allo-HSCT) was a protective prognostic factor. Conclusions The predominant relapse pattern in our center was early recurrence, with bone marrow being the main relapse site. Multivariate analysis revealed that very early relapse and failure to achieve CR2 were independent adverse prognostic factors (P<0.05). CART combined with Allo-HSCT significantly improved outcomes in children with relapsed ALL.