While RCTs still have a place in today’s cancer research arena, our increased understanding of disease means they are an outdated method of assessing therapies in some cancer settings.I strongly believe that the clinical trial landscape needs to evolve beyond RCTs so that we are less reliant on them for the regulatory approval of new medicines. In some situations, there is no standard of care and thus no comparator treatment. For example, it is difficult to perform phase III trials in small subgroups of patients such as those with rare cancers or paediatric patients, or to evaluate therapies in patients harbouring tumours with rare mutations, such as NTRK gene fusions.In other cases, a therapy may be so effective – particularly in patients positive for a molecular biomarker – that it would be unethical to perform a study in which patients are randomised to a clearly inferior therapy. This was the case when the first PD-1-targeted therapy, pembrolizumab, was granted approval by the US Food and Drug Administration (FDA) for the treatment of unresectable or metastatic melanoma based on phase I evidence alone. Further clinical development was considered unnecessary.There is a clear difference between the USA and Europe in relation to the type of evidence that is accepted to support an application for the regulatory approval of a medicine. Despite several trials with platform-based protocols – such as basket trials – being conducted in Europe, to date, none of these have provided evidence contributing to the regulatory approval of a medicine in Europe. In contrast, the FDA approved vemurafenib for the treatment of BRAF V600-mutant Erdheim-Chester disease and pembrolizumab for the treatment of deficient mismatch repair tumours based on data from these types of trials.Real-world data may be of interest to support trial-based evidence, but it is still early days and guidance is needed on how the oncology community can make the best use of such information.