The number of times that a certain procedure is performed in a hospital per year affects the outcome following this procedure. Such an effect on postoperative mortality has clearly been shown for many procedures in different countries. It is most outspoken for high-risk gastrointestinal surgery, e.g. pancreatectomy and oesophagectomy, and to a lesser degree for surgery for abdominal aneurysms. The experience of the surgical team, which includes surgeons, anaesthetists, nurses and paramedical staff, is responsible for one half of the beneficial effect and the surgeon for the other half. The percentage varies for different procedures, but teamwork and the hospital are important factors in all high-risk procedures. Pancreatectomies, oesophagectomies and high-risk vascular surgery should be concentrated in dedicated centres with a sufficient volume to obtain a good outcome of surgical care.