Background: Prophylactic use of rhG-CSF is routinely recommended in patients with a high risk of febrile neutropenia (FN) after chemotherapy. Currently, rhG-CSF is mostly used in clinical practice. However, PEG-rhG-CSF tends to be a substitute due to its convenience and long-acting effect. Regarding concerns about its efficacy and safety, whether PEG-rhG-CSF could be prophylactically used after these regimens with a high risk of FN is still in doubt. Doxorubicin combined with ifosfamide is commonly applied in sarcoma patients and is one of the most serious myelosuppressive chemotherapy regimens in solid tumors. We compared the efficacy and safety of PEG-rhG-CSF with rhG-CSF in the prevention of severe granulocytopenia and FN, in patients receiving this representative regimen. Methods: This was a single-center, prospective, randomized, crossover, non-inferiority phase II trial. Patients with sarcoma suitable for doxorubicin combined with ifosfamide treatment were enrolled. All patients received 2 cycles of chemotherapy and were randomly assigned to either the AOB group (PEG-rhG-CSF used in 1st cycle and rhG-CSF in 2nd cycle) or BOA group (rhG-CSF used in 1st cycle and PEG-rhG-CSF in 2nd cycle). Blood monitoring every other day until ANC ≥ 2×10⁹/L in two consecutive tests. The primary endpoints were the frequency and duration of grade 4 granulocytopenia after prophylactic use of PEG-rhG-CSF and rhG-CSF respectively, the secondary endpoints were the frequency and duration of febrile neutropenia, and proportion of antibiotic use. Results: From Jan 1, 2018 to Aug 31, 2021, 70 patients were enrolled, 35 each in AOB and BOA. There were no statistical differences in the incidence and duration of grade 4 granulocytopenia (85.7% vs. 77.0%, p = 0.1; 1.34d vs. 1.41d, p = 0.70), the proportion and duration of febrile neutropenia (12.8% vs. 8.2%, p = 0.77; 0.6d vs. 0.8d, p = 0.74), and the proportion of antibiotics used (19.1% vs. 14.9%, p = 0.79) when comparing PEG-rhG-CSF vs. rhG-CSF in prophylactical use. The incidence of grade 4 granulocytopenia in both groups was comparable in 1st cycle (90.9% vs. 96.0%, p = 0.47), but in 2nd cycle it was significantly lower in AOB than that in BOA (54.5% vs. 88.0%, p = 0.01). The lowest value of ANC was higher in AOB in 2nd cycle (0.32G/L vs. 0.12G/L, p = 0.03). Conclusions: This study demonstrated PEG-rhG-CSF could prevent granulocytopenia and FN in sarcoma patients receiving doxorubicin combined with ifosfamide, with comparable efficacy and toxicities as rhG-CSF. Considering the convenience in use and patient compliance, PEG-rhG-CSF could be an effective and reasonable alternative to rhG-CSF. Clinical trial information: ChiCTR1900021945.