Pancreatic cancer is associated with a very poor prognosis; however, in selected patients, resection may improve survival. Several recent reports have demonstrated that concentration of treatment activity for patients with pancreatic cancer has resulted in improved outcomes. The aim of this study was to ascertain if there was any evidence of benefit for specialised care of patients with pancreatic cancer in Scotland. Records of patients diagnosed with pancreatic cancer during the period 1993-1997 were identified. Three indicators of co-morbidity were calculated for each patient. Operative procedures were classified as resection, other surgery or biliary stent. Prior to analysis, consultants were assigned as specialist pancreatic surgeons, clinicians with an interest in pancreatic disease or nonspecialists. Data were analysed with regard to 30-day mortality and survival outcome. The final study population included 2794 patients. The 30-day mortality following resection was 8%, and hospital or consultant volume did not affect postoperative mortality. The 30-day mortality rate following palliative surgical operations was 20%, and consultants with higher case loads or with a specialist pancreatic practice had significantly fewer postoperative deaths (P=0.014 and 0.002, respectively). For patients undergoing potentially curative or palliative surgery, the adjusted hazard of death was higher in patients with advanced years, increased co-morbidity, metastatic disease, and was lower for those managed by a specialist (RHR 0.63, 95% CI 0.50-0.78) or by a clinician with an interest in pancreatic disease (RHR 0.63, 0.48-0.82). The risk of death 3 years after diagnosis of pancreatic cancer is higher among patients undergoing surgical intervention by nonspecialists. Specialisation and concentration of cancer care has major implications for the delivery of health services.