A prospective, randomized clinical trial was conducted to compare the efficacy of piperacillin/tazobactam and amikacin combination with carbapenem monotherapy for the empirical treatment of febrile neutropenic episodes of children with acute lymphoblastic leukemia or acute myeloblastic leukemia. Patients aged 2-16 years with hematological malignancies who had febrile neutropenia were randomly assigned to receive piperacillin/tazobactam (80 mg/kg piperacillin/10 mg/kg tazobactam, q6h) combined with amikacin (PTA) (7.5 mg/kg, q12h) or meropenem or imipenem (20 mg/kg, q8h) (C). Response to antimicrobial therapy, evaluated for etiological agents, was measured. Duration of fever, neutropenia, and hospitalization, mortality, and the need for additional antibiotics or antifungal drugs were compared for the treatment success between the two groups. Out of 87 febrile neutropenic episodes that were evaluable for comparison, 46 patients received PTA and 41 patients were treated with carbapenems (imipenem or meropenem). Overall, the microbiologically documented infection rate was 21.9%, with Staphylococcus epidermidis as the most common cause of bacteremia. The rate of treatment modification was 56.5% in the PTA group and 53.6% in the carbapenem group with no statistical difference (p > .05). There was no infection-related mortality during the study period. There was no difference between the two regimens for durations of fever, neutropenia, and hospitalization (p > .05 for all categories). PTA was as effective as carbapenem monotherapy as an initial empirical regimen in febrile neutropenic episodes of pediatric hematological malignancies.