PURPOSE: To compare the outcomes of paediatric and adolescent extracranial malignant germ cell tumour (GCT) patients treated with either carboplatin or cisplatin on clinical trials conducted by the Children's Oncology Group (COG) and the Children's Cancer and Leukaemia Group (CCLG). METHODS: The Malignant Germ Cell International Consortium (MaGIC) has created a database of the GCT clinical trials conducted since 1983 by COG (United States, Canada and Australia), which used cisplatin-based regimens, and by CCLG (United Kingdom), which used carboplatin-based regimens. Using the parametric cure model, this study compared the overall 4-year event-free survival (EFS), stratified by age, stage, site and the a-priori defined MaGIC 'risk' groups: standard risk ((SR) 1 (EFS >80%; age <11 years), SR2 (EFS >80%, age ≥ 11y) and poor risk (PR) (EFS ≤ 70%, age ≥ 11y). RESULTS: Cisplatin-based therapy was used in 620 patients; carboplatin was used in 163 patients. In the overall multivariate cure model, the two regimens did not differ significantly (cisplatin: 4-year EFS 86%; 95% confidence interval (CI) 83-89% versus carboplatin 4-year EFS 86%; 95% CI 79-90%; p = 0.87). No significant differences were noted in stratified analyses by site, stage, age and MaGIC risk groups: SR1 (p = 0.20), SR2 (p = 0.55) or PR (p = 0.72) patients. CONCLUSIONS: In these trials conducted contemporaneously, there is no significant difference in outcome observed overall, or any subset of patients, who were treated with regimens containing cisplatin versus carboplatin These results suggested sufficient equipoise to justify a randomised trial to evaluate the effectiveness of carboplatin versus cisplatin in the treatment of children, adolescents and young adults with standard risk GCT, which is currently underway.