High morbidity and mortality rates are reported for bypass and resective surgery of pancreatic cancer. In a retrospective study we correctly predicted the postoperative course in 88% of the patients who underwent bypass surgery and 83% of those who had a resection for pancreatic cancer. Before starting with clinical application of this scoring system, we undertook a prospective study to confirm its predictive value. Sixty-seven consecutive patients with pancreatic cancer were included: 42 patients underwent bypass surgery and 25 pancreatic resections. The operative mortality was 14% for palliative surgery and 0% for resective surgery. Surgical team and nurses were totally unaware of the predicted risk. The preoperative forecast proved to be correct in 81% of bypass surgery and in 88% of resective surgery, although surgical mortality had decreased from 21 to 14% for bypass surgery and from 17 to 0% for resective surgery. Tables are included to calculate the surgical risk for each of 162 combinations of the risk factors considered in the predictive model (81 for bypass surgery and 81 for resective surgery). Calculation of surgical risk is important when evaluating different treatments for pancreatic cancer are available.