This meeting, organised in association with the working party on arrhythmias of the European Society of Cardiology, provided one-and-a-half days of fruitful interchange between workers from various European countries, the United States, Canada, and China. An essential prerequisite of the proper management of arrhythmia is the careful understanding of mechanisms, incidence and specific features, which was provided by the first half day. It is important to appreciate the incidence of arrhythmias in the apparently healthy population, and to compare it with that seen in patients. In addition to the conventional electrocardiographic tracing we now, from many centres, have further information obtained by long-term ambulatory monitoring. From these data we can correlate aspects of biological variability in arrhythmias, including the influence of the autonomic nervous system. Delayed conduction is often an important marker of a tendency to develop arrhythmias and the possibility of recording late potentials from the surface of the body is a great step forward from having always to depend on invasive investigations. Nevertheless, the precise determination of the nature of certain arrhythmias and the assessment of possible therapeutic benefit still depends on programmed electric stimulation of the heart with appropriate recordings. We are thus usefully reminded of the fact that different information may be obtained from different sites. While we would like to be able to tailor antiarrhythmic therapy precisely to the type of rhythm disorder that has been recorded, this has not yet been achieved with anything like the degree of confidence that one knows from other clinical situations e.g. antibacterial therapy. Indeed, we still need to agree on the definition of high risk regarding the development of arrhythmias including those likely to cause sudden death, whether in the context of ischaemic or other heart disease.