Pancreatic cancer has a poor prognosis with complete surgical resection being the only therapy to offer a realistic chance for long-term survival. The aim of this study is to identify surgery-related variables that influence long-term survival. Between 1990 and 2002, 226 consecutive patients (mean age of 64+/-11 years) had resection for pancreatic adenocarcinoma. Prognostic variables in these patients were analyzed using univariate and multivariate analysis. Two hundred four patients (90%) had pancreaticoduodenectomy, 13 patients (6%) had distal pancreatectomy, and 9 patients (4%) had a TP. Stage I disease was present in 50 (22%), stage II disease in 170 (75%), and stage III disease in 6 (3%). R0 resections were achieved in 70%. Operative morbidity was 36% and 30-day mortality was 6%. Actual 1-year, 3-year, and 5-year survival rates were 49% (n=111), 14% (n=31), and 4% (n=9). Using multivariate analysis: tumor size, tumor differentiation, obtaining an R0 resection, and lack of postoperative complications were variables associated with long-term survival. Long-term survival in patients with pancreatic cancer after resection remains poor. Achieving a margin negative resection (R0) with no postoperative complications are prognostic variables that can be affected by the surgeon.