With improvements in treatment for childhood cancer, comparisons of survival rates between countries have become important to inform future health policies and treatment strategies. Population-based cancer registry data are viewed as the gold standard for such comparisons, but even these have potential confounding factors. Here, we review the interpretation of recent geographical comparisons of childhood cancer survival from the viewpoint of the British Isles, a region with a 45-year record of national population-based cancer registration and a national childhood cancer clinical trials organisation in place for nearly 30 years. Using national data on referral patterns to tertiary paediatric oncology centres, we explore some of the reasons for lower survival rates in the past for some tumour groups and anticipate continued improvement in the next decade. Participation in international clinical trials coincided with rapid gains in survival for hepatoblastoma. This exemplifies the potential benefits of international collaborative clinical research, particularly for rare subgroups.