In an evidence-based review of the relationship between volume and quality of care, the independent Dutch Institute for Healthcare Improvement (CBO) concluded that volume appears to be related to outcome for certain surgical procedures (case fatality after pancreatic and oesophageal cancer) and that quality of care might be improved by centralisation. The Dutch Institute for Healthcare Improvement also identified conditions required for centralisation, particularly acceptance by professionals and hospitals. In the USA, programmes to improve quality ofcare initiated by the Leapfrog Group using volume criteria or, more recently, using 'public reporting' and 'pay for performance' principles have led to improvements in quality. In Canada, the Surgical Oncology programme within the Cancer System Quality Index programme has reduced case fatality following pancreatic resection. The Canadian programme was based not only on volume but also on standards, guidelines, rapid access strategies and publicly available performance assessments. In The Netherlands, the Dutch Health Care Inspectorate is introducing the so-called performance indicators of care. Other initiatives are underway to develop a system with multiple quality criteria as in Canada. These programmes should not be restricted to surgical procedures but should include complex procedures in other specialties as well.