Objective: To analyze the treatment-free remission (TFR) outcomes after discontinuation of tyrosine kinase inhibitor (TKI) in children with chronic myeloid leukemia (CML). Methods: In this retrospective cohort study, clinical data of 14 chronic phase CML children aged <18 years who had achieved stable deep molecular response (DMR) for ≥ 2 years after standardized treatment with TKI and had a strong desire to discontinue TKI at Henan Cancer Hospital from September 30, 2016 to January 30, 2022 were collected retrospectively. According to the different TFR outcomes after discontinuation of TKI, patients were divided into loss of major molecular response (MMR) group and without loss of MMR group, differences in clinical characteristics between the two groups of children were analyzed using Mann-Whitney U test and Fisher exact test. Results: Out of 14 children with TKI discontinuation, 7 were male and 7 were female. The age at diagnosis was 14.0 (4.8, 17.0) years, and the age at TKI discontinuation was 22.0 (12.5, 27.0) years. Among them, 8 children were treated with imatinib prior to TKI discontinuation and 6 children were treated with second-line substitution of the second-generation TKI nilotinib or dasatinib prior to TKI discontinuation. The follow-up time was 37.0 (27.8, 47.5) months, and 7 cases lost MMR at the time of discontinuation of 3.0 (2.0, 11.0) months. Eight children gained TFR at 6 months, 7 children gained TFR at 12 and 24 months. Amongst the 6 children who received second-generation TKI prior to TKI discontinuation, 2 children lost MMR at 3 and 11 months and 4 children gained TFR, among the 8 children who discontinued imatinib, 5 children lost MMR at the time 3.0 (2.0, 9.0) months and 3 children gained TFR. The age at diagnosis and TKI discontinuation, the time from TKI treatment to the acquisition of DMR, the duration of TKI treatment before TKI discontinuation, the duration of DMR before TKI discontinuation, and the number of children treated with second-generation TKI were not statistically different between the 7 children in the group that did not lose the MMR and the 7 children in the group that lost the MMR (all P>0.05). All the 7 children with confirmed loss of MMR immediately restarted TKI therapy, and all regained DMR after 2.0 (2.0, 11.0) months of therapy. None of the children had disease progression. After TKI discontinued, only 1 child had mild bone pain, which could be relieved by oral antipyretic analgesic drugs. Conclusions: Children with CML who have achieved a durable stable DMR for≥2 years on TKI therapy can discontinue the TKI and obtain TFR. Both the longer duration of TKI therapy, the longer duration of DMR and the use of second-generation TKI therapy before TKI discontinuation, may allow more children with CML who are expecting TKI discontinuation to have access to TFR.