Abstract Background Acute lymphoblastic leukemia (ALL) in children currently has more than 90% survival in higher resource countries. However, in less resource countries, the survival rates are lower and infections are frequent causes. Risks for infections and poor outcomes vary during the chemotherapy phases. There is a paucity of data concerning the frequencies of the chemotherapy phase where the event is more probable. Our primary objective is to describe infections frequencies and outcomes of infection in children and also to look for association with risk factors. Methods Patients (Pts) were treated according to ALLIC-BFM-2010 protocol, a simpler branch of the BFM-AEIOP protocol and accrued prospectively. The main question is comparing outcomes of two randomizations using higher dose chemotherapy based on the leukemia risk. The risk severity was based on levels of measurable residual disease (<0.1, 0.1-10, >10 %,) and categorized in high (HR), intermediate (IR), and low (LR) risks. In SR Pts with B-cell-precursor ALL two doses of daunorubicin were prescribed in induction compared with four doses in all others. In consolidation, high-dose methotrexate was administered at 5g/m2 for SR/IR T-cell ALL and at 2g/m2 for SR/IR B-precursor-ALL. Consolidation for HR pts consisted of three intensive polychemotherapy blocks. Only T-ALL and HR pts age > 1 year received prophylactic cranial radiotherapy (12 Gy). Therapeutic cranial radiotherapy was reserved for pts with initial CNS. We performed a descriptive analysis, comparison of frequencies and proportions using Fisher test/Ji2. For time to event proportion we used Cox Analysis. Results 2232 de novo ALL pts were analyzed. Median age 5y (IQR 3-10), 56.3% were males. 2.1% had Down Syndrome; 24% categorized as HR, 56% as IR, and 20% as LR. The Immunophenotype analysis mostly B lineage (88.8%) and less T lineage (11.2%). The majority (96.7%) of pts achieved complete remission. Estimated mean time of survival was 8.8 years (ES 0.099, CI 95% 8.38-8.77) and the overall survival rate was 74 % (0.01). Trial wide, 518 (23.2%) pts died, 59 (2.6%) pts had criteria for clinical sepsis and 459 (20%) for not related sepsis causes. According to phase of the protocol, in less than one month from diagnosis, 21 events (35.6%) were observed; from 2-3 months 9 events (15.3%), from 4-6 months 6 events, (10.2%) and afterwards 23 events (39%) (p-value 0.000). No difference was found on T vs B immunophenotype, sex, age or Down syndrome. (P-value NS). However significant differences were found on risk groups 26/538 pts (4.8%) died on HR, 24/1248 (1.9%) in IR and 9/433 (2.1%) on standard risk (p-value 0.009). There was an association between the rates of sepsis and death on pts randomized to I and II protocols (p-value 0.007 and P-value 0.007). Conclusion In this large cohort of pts, we could demonstrate association between sepsis and mortality in those high-risk groups and in those pts randomized to higher chemotherapy doses. These results will allow to establish strategies to prevent infections or maximize medical interventions preventing poor infectious outcomes in high-risk groups.