Optimal treatment for unresectable carcinoma of the pancreas remains controversial. This study was done to examine the relationship between perioperative jaundice and postoperative morbidity, and type of palliative biliary bypass and postoperative morbidity and jaundice clearance. Seventy-six patients with obstructive jaundice secondary to carcinoma of the head of the pancreas were studied. Forty-nine patients underwent one of four different types of palliative bypass: 1, cholecystojejunostomy (n = 22); 2, choledochojejunostomy (n = 11); 3, choledochoduodenostomy (n = 9), and 4, cholecystoduodenostomy (n = 7). Age, sex and preoperative health status were similar for all operative groups, as well as for those with and without postoperative morbidity. The postoperative complication rate was 33 per cent and there was one postoperative death. Length of preoperative jaundice and peak preoperative bilirubin levels were independent of morbidity. Postoperative morbidity was similar for each type of bypass used and no significant difference was found when cholecystoenteric (1 and 4) and choledochoenteric (2 and 3) bypass were compared. The results of this study support the view that postoperative morbidity is not directly related to the presence of jaundice preoperatively. Furthermore, the rate of jaundice clearance and the occurrence of postoperative complications are not dependent on the type of bypass used.Optimal treatment for unresectable carcinoma of the pancreas remains controversial. This study was done to examine the relationship between perioperative jaundice and postoperative morbidity, and type of palliative biliary bypass and postoperative morbidity and jaundice clearance. Seventy-six patients with obstructive jaundice secondary to carcinoma of the head of the pancreas were studied. Forty-nine patients underwent one of four different types of palliative bypass: 1, cholecystojejunostomy (n = 22); 2, choledochojejunostomy (n = 11); 3, choledochoduodenostomy (n = 9), and 4, cholecystoduodenostomy (n = 7). Age, sex and preoperative health status were similar for all operative groups, as well as for those with and without postoperative morbidity. The postoperative complication rate was 33 per cent and there was one postoperative death. Length of preoperative jaundice and peak preoperative bilirubin levels were independent of morbidity. Postoperative morbidity was similar for each type of bypass used and no significant difference was found when cholecystoenteric (1 and 4) and choledochoenteric (2 and 3) bypass were compared. The results of this study support the view that postoperative morbidity is not directly related to the presence of jaundice preoperatively. Furthermore, the rate of jaundice clearance and the occurrence of postoperative complications are not dependent on the type of bypass used.