PURPOSE Our previous NECO phase II studies on high-grade osteosarcoma suggested that administering ifosfamide (IF; 16 g/m 2 [4g/m 2 once on day 1, then 2g/m 2 once on days 2-7] × six) to patients showing a poor response (PrRsp) to preoperative chemotherapy with methotrexate, doxorubicin, and cisplatin (MAP) improves their prognoses. In this Japan Clinical Oncology Group (JCOG) study, JCOG0905, we aimed to investigate the efficacy and safety of IF in patients with PrRsp. METHODS JCOG0905 is a multicenter, open-label, multi-institutional, randomized trial. Eligible patients (50 years and younger) had resectable, high-grade osteosarcoma (stage II or III, Union for International Cancer Control TNM) of the extremities, limb girdles, and thoracic wall. After two MAP cycles and tumor resection, patients with PrRsp were randomly assigned to either the MAP or MAP plus 15 g/m 2 (3g/m 2 once daily on days 1-5) × six IF (MAP + IF [MAPIF]) group. The primary end point was disease-free survival (DFS); secondary end points were overall survival (OS) and safety. The planned sample size was 100 patients with a one-sided α of .1 and a power of 0.7, assuming a 3-year DFS of 50% and 65% for MAP and MAPIF, respectively. This trial is registered with the Japan Registry of Clinical Trials (jRCT; jRCTs031180126). RESULTS Of the 287 patients registered between February 2010 and August 2020, 51 and 52 patients with PrRsp were assigned to the MAP and MAPIF groups, respectively. As of March 2022, DFS did not differ between groups (hazard ratio [HR], 1.05 [95% CI, 0.55 to 1.98]) and OS was numerically inferior in the MAPIF group (HR, 1.48 [95% CI, 0.68 to 3.22]). Nine and zero patients in the MAPIF and MAP groups discontinued treatment because of adverse events, respectively. CONCLUSION Evidence from JCOG0905 does not support the addition of IF for patients with PrRsp.