To summarize the clinical characteristics of high-risk neuroblastoma (HR-NB) in a single center, analyze the prognostic factors of HR-NB. The clinical data of children with HR-NB who were treated and followed up at the hematology-oncology center of Beijing Children’s Hospital from February 1, 2007 to June 30, 2018 were analyzed retrospectively. The clinical features were summarized. Kaplan-Meier method was used for survival analysis and Cox regression was used to analyze the prognostic factors. The last follow-up time was June 30, 2019. A total of 458 children with HR-NB were enrolled in this study, including 265 males (57.9%) and 193 females (42.1%), the age at diagnosis was 40.0 months (4.5-148.0 months), the follow-up time was 22.0 months (0.2-138.0 months) and the time of tumor progression or recurrence was 15 months (1-72 months). The 5-year event-free survival (EFS) rate was (31.2±2.6)% and the 5-year overall survival (OS) rate was (43.9±3.2)%. The 5-year EFS rate and 5-year OS rate in 142 hematopoietic stem cell transplantation (HSCT) patients with bone marrow metastases were better than that in 196 non-transplantation cases with bone marrow metastases ((26.5±4.5)% . (25.1±3.6)%, χ²=13.773, 0.001; (38.1±5.5)% . (35.7±4.7)%, χ²=9.235, 0.002); 128 transplantation patients with bone metastases had higher 5-year EFS rate and 5-year OS rate than 188 non-transplantation cases with bone metastases ((28.5±5.0)% . (26.7±3.8)%, χ²=10.222, 0.001; (37.1±6.0)% . (36.2±4.8)%, χ²=7.843, 0.005). The 5-year EFS rate was higher in 37 HSCT patients with MYCN amplification than in 49 non-transplantation cases with MYCN amplification ((26.8±8.0) % . (20.5±6.4) %, χ²=5.732, 0.017). No significant difference was found in 5-years OS rate between transplantation group with MYCN amplification and non-transplantation group with MYCN amplification ((31.4±8.6) % . (26.2±7.4) %, χ²=3.230, 0.072). Univariate survival analysis showed that lactate dehydrogenase (LDH)≥1 500 U/L was associated with poor prognosis of patients with MYCN amplification (χ²=6.960, 0.008). Multivariate Cox analysis showed bone marrow metastasis and LDH≥1 500 U/L were independent risk factors for poor prognosis of patients with non-MYCN amplification (=2.427, 1.618;95：1.427-4.126, 1.275-2.054, 0.05) for both comparisons. LDH≥1 500 U/L was the poor prognostic factor for patients with MYCN amplification. The bone marrow metastasis and LDH≥1 500 U/L were the poor prognostic factors for HR-NB patients with non-MYCN amplification. HSCT can improve the prognosis of patients with bone or bone marrow metastasis. It can also retard the time of progression or recurrence for patients with MYCN amplification.