Both patients and government demand proof of quality of care and value for money. Our unit ist in a large teaching hospital, performing over 6000 procedures per year. We have designed and implemented a Quality Assurance (QA) program, the basis of which we believe could be a model for endoscopy QA. A QA team was formed and a literature search undertaken. An initial 3 month audit was then performed into indications for, and complications of, all procedures. The results of this initial audit led us to concentrate on colonoscopy and ERCP. The specific items of data collected were based on the "Core Quality Indicators" developed by the American Society for Gastrointestinal Endoscopy (ASGE). We also analysed data relating to endoscope disinfection, equipment failure and carried out a patient satisfaction survey. The data were presented at 3-monthly QA meeting, and appropriate action taken. We performed a detailed audit of ERCP (217 procedures) and colonoscopy (904 procedures). Patients risk was stratified using the American Society of Anaesthesiology (ASA) classification. Using these data we established our technical success and complication rates for colonoscopy and ERCP. Audit of equipment revealed that on average an endoscope was away for repair 9 % of the time. Contamination of endoscopes was frequent with glutaraldehyde disinfection; the rate of contamination fell dramatically when we changed our disinfection method. A QA program can be implemented in busy endoscopy units. There are significant problems, however, in ensuring that such a program is effective: these include inadequate funding/staffing, lack of suitable information technology and lack of clear guidelines for dealing with poor performance.