To reduce the incidence of febrile neutropenia during rapid COJEC (cisplatin, vincristine, carboplatin, etoposide, and cyclophosphamide given in a rapid delivery schedule) induction. In the High-Risk Neuroblastoma-1 (HR-NBL1) trial, the International Society of Paediatric Oncology European Neuroblastoma Group (SIOPEN) randomly assigned patients to primary prophylactic (PP) versus symptom-triggered granulocyte colony-stimulating factor (GCSF; filgrastim). From May 2002 to November 2005, 239 patients in 16 countries were randomly assigned to receive or not receive PPGCSF. There were 144 boys with a median age of 3.1 years (range, 1 to 17 years) of whom 217 had International Neuroblastoma Staging System (INSS) stage 4 and 22 had stage 2 or 3 MYCN-amplified disease. The prophylactic arm received a single daily dose of 5 microg/kg GCSF, starting after each of the eight COJEC chemotherapy cycles and stopping 24 hours before the next cycle. Chemotherapy was administered every 10 days regardless of hematologic recovery, provided that infection was controlled. The PPGCSF arm had significantly fewer febrile neutropenic episodes (P = .002), days with fever (P = .004), hospital days (P = .017), and antibiotic days (P = .001). Reported Common Toxicity Criteria (CTC) graded toxicity was also significantly reduced: infections per cycle (P = .002), fever (P < .001), severe leucopenia (P < .001), neutropenia (P < .001), mucositis (P = .002), nausea/vomiting (P = .045), and constipation (P = .008). Severe weight loss was reduced significantly by 50% (P = .013). Protocol compliance with the rapid induction schedule was also significantly better in the PPGCSF arm shown by shorter time to completion (P = .005). PPGCSF did not adversely affect response rates or success of peripheral-blood stem-cell harvest. Following these results, PPG-GSF was advised for all patients on rapid COJEC induction.