The endoscopic approach to biliary drainage came late on the scene; some of the published results reflect early experience with inadequate techniques. Now it is clear that the endoscopic approach is preferable to the percutaneous method. When palliation of jaundice is required (in a patient without impending duodenal obstruction), there is a simple choice between surgical bypass or endoscopic stenting. Stenting is substantially cheaper than surgery--at least for the initial admission. Recovery from stenting is almost immediate, which cannot be said for surgical intervention. Time will tell how far the need for readmission (stent blockage, duodenal obstruction) will erode these advantages. The main factor influencing our decision (stent or surgery), apart from the hope of resection, is the patient's general status, or "operative risk." Unfortunately, there is no accepted risk factor scale or template against which our experiences can be compared. There are no absolutes, only a spectrum of patients who differ according to the tumor load and their general medical condition. A fit patient with a relatively small tumor is best served by surgical intervention. The diagnosis and its unresectable nature can be established beyond doubt, and anastomoses (biliary and gastroduodenal) can be established of such a size that subsequent obstruction is unlikely. The operative mortality rate will be low. Patients with a large tumor load and poor general condition are best served by an endoscopic stent. Between these positions lies a spectrum of patients and plenty of room for discussion and personal opinions. Drainage procedures are unwarranted in patients who are truly terminal. Specialist vested interests have seriously jaundiced the view of many people in this field--and obstructed attempts at consensus. As in the management of patients with gallstone disease, it is important that surgeons, endoscopists, and radiologists work together as teams in the best interests of our patients, present and future.