Objective: To analyze the short-term clinical efficacy and safety of arsenic trioxide (ATO) combined with a modified N7 induction regimen in the treatment of children with high-risk neuroblastoma (NB). Methods: This study was a prospective, single-arm, multicenter phase Ⅱ clinical study. Sixty-seven high-risk NB children from eight units of Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Wuhan Children's Hospital of Tongji Medical College of Huazhong University of Science and Technology, First Affiliated Hospital of Guangxi Medical University, Hainan General Hospital, Affiliated Hospital of Guangdong Medical University, Kunming Children's Hospital, Tongji Hospital of Tongji Medical College of Huazhong University of Science and Technology, and Guangdong Provincial Agricultural Reclamation Center Hospital were enrolled from January 2019 to August 2023 and were treated with ATO combined with a modified N7 induction regimen. The efficacy and adverse effects at the end of induction chemotherapy were assessed and analyzed, and the differences in the clinical characteristics were further compared between the treatment-responsive and treatment-unresponsive groups by using the Fisher's exact test. Results: Among 67 high-risk NB children, there were 40 males (60%) and 27 females (40%), with the age of disease onset of 3.5 (2.6, 4.8) years. Primary NB sites were mostly in retroperitoneum (including adrenal gland) (56/67, 84%) and the common metastases sites at initial diagnosis were distant lymph node in 25 cases (37%),bone in 48 cases (72%),bone marrow in 56 cases (84%) and intracalvarium in 3 cases (4%). MYCN gene amplification were detected in 28 cases (42%). At the end of induction, 33 cases (49%) achieved complete remission, 29 cases (43%) achieved partial remission, 1 case (1%) with stable disease, and 4 cases (6%) were assessed as progressive disease (PD). The objective remission rate was 93% (62/67) and the disease control rate was 94% (63/67). The percentage of central system metastases at the initial diagnosis was higher in the treatment-unresponsive group than in the treatment-responsive group (2/5 vs. 2% (1/62), P=0.013), whereas the difference in MYCN gene amplification was not statistically significant between two groups (3/5 vs.40% (25/62), P=0.786). Grade Ⅲ or higher adverse reactions during the induction chemotherapy period were myelosuppression occurred in 60 cases (90%), gastrointestinal symptoms occurred in 33 cases (49%), infections occurred in 20 cases (30%), hepatotoxicity occurred in 4 cases (6%), and cardiovascular toxicity occurred in 1 case (2%). There were no chemotherapy-related deaths. Conclusion: ATO combined with N7-modified induction regimen had a superiority in efficacy and safety, which deserved further promotion in clinical practice.