Background: The prognosis of childhood acute lymphoblastic leukemia (ALL) in Indonesia, a lower-middle-income country (LMIC), is lower than in high income countries (HICs). The Indonesian ALL2013 protocol resulted in too many toxic deaths (21%) and abandonments (11%). Therefore, we drafted an adapted protocol, ALL2016. Main changes: no anthracyclines in standard risk (SR), prednisone replaced dexamethasone at induction in high risk (HR), and anthracyclines and cyclophosphamide were rescheduled in HR. Procedure: Patients (aged: 1–18 years) were stratified into SR and HR. HR was defined as age over 10 years, leucocyte count over 50 × 109/L, central nervous system (CNS) involvement, mediastinal mass, T-cell phenotype, testicular involvement, or poor prednisone response. Results: ALL2013 included 174 patients (106 SR and 68 HR) and ALL2016 188 (91 SR and 97 HR). Although the number of HR patients was significantly higher in ALL2016 (51.6% vs. 39.1%; p =.017), the outcome of ALL2016 improved over ALL2013 (4-year-probable overall survival (pOS) 60.1% vs. 50.0%; p =.042 and 4-year-probable event-free survival (pEFS) 49.5% vs. 36.8%; p =.018). ALL2016 showed a nonsignificant advantage for SR patients (4-year-pEFS 56.0% vs. 47.2%; p =.220 and 4-year-pOS 70.3% vs. 61.3%; p =.166), but less toxic deaths (7% vs. 20%; p =.011). In HR group, the outcomes were significantly better in ALL2016 (4-year-pEFS 43.3% vs. 20.6%; p =.004; 4-year-pOS 50.5% vs. 32.4%; p =.014) especially due to less relapses (31% vs. 62%; p =.001). Isolated CNS relapses went down from 18 to 8% in HR (p =.010) and 11 to 5% in SR (p =.474). Both SR and HR showed lower numbers of abandonment in ALL2016 (6% vs. 14%; p =.039). Conclusions: Overall ALL2016 results improved over ALL2013. Modest changes in protocol resulted in less initial toxicity and abandonments.