Abstract： Background Nowadays, there is a remarkable rise in resectability rate of periampullary adenocarcinoma and the mortality and morbidity of the pancreaticoduodenectomy procedure have been reduced remarkably, while the 5 year survival rates of patients with carcinoma of the head of the pancreas are still below 25%. We conducted this retrospective study to evaluate the clinical outcome of radical pancreaticoduodenectomy with extended retroperitoneal lymphadenectomy as a surgical therapy for adenocarcinoma of the head of the pancreas.Methods Twenty cases with adenocarcinoma of the head of the pancreas were treated by standard pancreaticoduodenectomy (removing only the peripancreatic lymph nodes en bloc with the tumour) from 1994 to 1997, and 46 cases with the same disease underwent extended retroperitoneal lymphadenectomy associated with standard pancreaticoduodenectomy from 1998 to 2002. The patients for whom there were insufficient follow-up data, or who had received postoperative adjuvant therapy, were excluded from the analysis. Clinical and pathological parameters of both groups were reviewed. The postoperative morbidity, mortality and survival data were compared statistically. Results Demographic and histopathological characteristics were similar in the two groups of patients. Performance of the extended lymphadenectomy lengthened the procedure. The mean total number of lymph nodes resected was significantly higher in the radical group (P< 0.05). Of the 46 cases in the radical group, 26% (12/46) had metastatic adenocarcinoma in the resected retroperitoneal lymph nodes. There was one perioperative death in the standard group, and two in the radical group. Postoperative diarrhoea and lymphatic leakage were only observed in the radical group. Transfusion requirements and postoperative morbidity rates did not differ between the two groups. The 1-, 2- and 3-year survival rates were 63%, 32% and 21% respectively in the standard group, and 66%, 38% and 21% in the radical group. No statistically significant difference was found between the groups. When subgroups of node positive patients were analysed, the 1-, 2- and 3-year survival rates were 42%, 17% and 8% respectively in the standard group, and 65%, 32% and 16% in the radical group. Better survival was observed in the first 2 years after operation in the radical group, but no survival differences were seen after 2 years post operation. Conclusions The addition of an extended lymphadenectomy to a pancreaticoduodenectomy did not significantly increase morbidity rates, but was associated with an early survival advantage.