The steep pressure gradient that occurs at the plasma edge is the best known example of a transport barrier in magnetically confined plasmas, giving rise to improved confinement: the H-mode. However, most H-modes experience some form of repetitive relaxation of the barrier by what are called Edge Localised Modes (ELMs). These ELMs occur in apparently different types, and the physics of the instabilities that give rise to them is an important and interesting plasma research topic.Edge localised modes with their periodic ejection of particles and energy from the plasma edge are a characteristic of standard H-mode operation which is the baseline regime for ITER. It is the potential for ELMs to damage plasma facing components (PFCs) in a next-step tokamak which has focused effort in this area and the implications of ELMs for the ITER divertor is the subject of our first paper in this cluster (Federici et al). Current devices experience no significant direct damage to PFCs due to ELMs but extrapolations to ITER based on a multi-machine database suggest that ELM size will be comparable to the acceptable limits (Loarte et al). Detailed investigation of parametric dependencies of ELM energy and particle losses in specific machines can test the dependencies seen in the multi-machine approach against a data set possessing the maximum consistency and reveal other important details (Urano et al).One of the ideas to come out of the multi-machine study of ELM size is the possibility that parallel transport in the scrape-off-layer (SOL) determines ELM size. The response of the SOL to ELMs is reported in detail for DIII-D, revealing that some observations fit with this simple model while others do not (Fenstermacher et al). Radial transport during ELMs is also important both for understanding the physics of ELM events and also for predicting the magnitude of ELM interaction with the main chamber walls in future tokamaks (Gonçalves et al).It is clear that control of ELM frequency and size will be necessary in burning plasma experiments. Experimental results from the TCV tokamak are reported here which demonstrate triggering of ELMs by rapid vertical movements of the plasma (Degeling et al). In JET the use of seeded impurities to control type I ELM size and frequency whilst maintaining high confinement and density has been explored (Maddison et al).ELMs result from a transient collapse of the edge transport barrier (ETB), and the theoretical understanding of the ideal magnetohydrodynamic (MHD) stability limits for this region is therefore critical for understanding ELM phenomenology and scaling. Calculations of ballooning and kink stability limits at intermediate to high toroidal mode number are now highly advanced and indicate the important role of edge current and plasma shape (Snyder and Wilson). The ultimate goal of ELM modelling is an integration of edge stability codes with core transport and edge fluid codes to provide a complete predictive capability for ELMs. The final paper in this cluster shows how such an approach can be applied to interpretation of experimental data from JET (Lönnroth et al).Guy Matthews