Background: The addition of tyrosine kinase inhibitors (TKIs) to chemotherapy has dramatically improved outcomes of patients with Philadelphia chromosome (Ph)-positive acute lymphoblastic leukemia (ALL). When allogeneic hematopoietic stem cell transplant (HSCT) is performed, maintenance TKI is generally given for a fixed duration. However, the optimal duration of TKI outside of HSCT remains unknown, and the common practice is to continue indefinitely. Here, we report characteristics and outcomes of 9 patients treated with chemotherapy + TKI without HSCT and later discontinued TKI. Methods: Among 188 patients with Ph-positive ALL who did not undergo HSCT, 9 of them discontinued maintenance TKI mainly due to side effects. Patients were closely monitored with serial PCR testing for the BCR-ABL1 transcript. Major molecular response (MMR) was defined as BCR-ABL1 transcript ≤0.1% on the international scale for p210 transcripts and a 3-log reduction from baseline for p190 transcripts. Deep molecular remission (DMR) was defined as the absence of quantifiable BCR-ABL1 transcripts with a sensitivity of 0.01%. Molecular relapse was defined as loss of MMR. Treatment-free remission (TFR) was defined from time of TKI discontinuation to molecular relapse, last follow-up, or death from any cause. Results: At the time of TKI discontinuation, transcript level was undetected in 6 patients, <0.01% in 2 patients, and 0.01% in another patient. Prior to discontinuation, the median duration of TKI therapy and of DMR was 70 and 47 months, respectively. No morphological relapse occurred. Three patients (33%) had molecular relapse at a median of 6 months. All 3 resumed TKI therapy, and 2 of them regained DMR after a median of 13 months. After a median follow-up of 49 months, the median TFR was not reached, and the 4-year TFR rate was 65%. The median duration of DMR in patients with and without molecular relapse was 22 and 58 months, respectively (p = 0.096). Conclusion: TKI discontinuation outside of HSCT in Ph-positive ALL in the setting of compelling toxicity may be safe only among a highly selected group of patients with deep and prolonged DMR undergoing close and frequent monitoring. Validation of these findings in prospective clinical trials is highly needed.