It has been suggested that the placement of endoscopic or percutaneous biliary stents prior to pancreaticoduodenectomy increases postoperative morbidity. A retrospective review of a prospectively collected database was performed. Patients undergoing preoperative biliary stenting were compared with patients who did not undergo stenting. In addition, outcomes after endoscopic and percutaneous stenting were compared. Patients who had undergone operative biliary bypass prior to pancreaticoduodenectomy were excluded from the analysis. Between January 1994 and December 1997, 567 patients underwent pancreaticoduodenectomy without prior operative biliary bypass. Preoperative biliary stenting was performed in 408 patients (72%), whereas the remaining 159 patients (28%) did not undergo biliary stenting. In the stented group, 64% had stents placed via a percutaneous approach and 36% had stents placed endoscopically. The stented patients were older (mean 63.1 years vs. 61.4 years; P = 0.05) and were more likely to be white (92% vs. 82%; P = 0.005). Those who had stents placed were more likely to have jaundice (67% vs. 38%; P <0.0001) and fever (5% vs. 1%; P = 0.03) as presenting symptoms. There were no differences in multiple intraoperative parameters when the two groups were compared. Patients who had stents placed had a perioperative mortality rate of 1.7% compared to 2.5% in those who did not (P = 0.3). Although the overall complication rates were 35% in those who had stents placed and 30% in those who did not (P = NS), patients with stents experienced a significantly increased incidence of pancreatic fistula (10% vs. 4%; P = 0.02) and wound infection (10% vs. 4%; P = 0.02). The incidences of other postoperative complications were similar between the stented and unstented groups. Eight patients (3%) in the percutaneously stented group developed significant hemobilia after stent placement, whereas none of the patients undergoing endoscopic stent placement developed hemobilia (P = 0.03). There were no statistical differences in other complications between the percutaneously and endoscopically stented groups. Preoperative biliary stenting did not increase the overall complication rate or mortality rate in patients undergoing pancreaticoduodenectomy. Stenting does appear to increase the rate of pancreatic fistula formation, possibly as a result of pancreatic inflammation related to the stenting procedure. Stenting also increases the rate of wound infection, likely secondary to contaminated bile (bactibilia) after instrumentation of the biliary tree. Preoperative biliary stenting is safe but should be used selectively because of the above-mentioned risks. The method of stenting should be based on local expertise.