Objective To explore the predictive value of minimal residual disease (MRD) level at different time-points of treatment in children with precursor B- acute lymphoblastic leukemia (B-ALL) in the context of MRD-guided therapy. Methods Data of newly diagnosed 417 patients with B-ALL from September 2014 to November 2017 were reviewed. We used multiparametric flow cytometry to monitor the MRD level on day 15, 33, 90 and 180, and analyzed the relationship between MRD levels and prognosis. Results The 417 patients included 240 males and 177 females with a median age of 5-years-old (3.0 year to 10.0 years). With a median follow-up of 44.0(33.7-56.2) months, the 3-y overall survival (OS) and event-free survival (EFS) were (90.9±1.4)% and (85.2±1.7)%, respectively. Patients who reached good MRD level on day 15 ((<10.0% or ≥10.0%), day 33 (<0.1% or ≥0.1%), day 90/180 (<0.01% or ≥0.01%) had a significantly higher probability of estimated OS and EFS (P<0.05). Patients who reached the MRD negative at all 3 time-point (day 33, 90, 180) had a significantly higher probability of estimated OS (95.8% vs. 82.8%) and EFS (92.3% vs. 72.6%) compared to patients with MRD failure at least in one time-point (P<0.05). Multivariable analysis showed MRD≥0.1% on day 33 and MRD≥0.01% on day 180 were independent risk factors for OS and EFS (P<0.05). Conclusion MRD levels on day 33 and 180 have important prognostic implications even in the context of MRD guided therapy. Sequential MRD monitoring is warranted in the treatment of children with B-ALL.