Between 1969 and 1986, 88 patients had a Whipple resection for adenocarcinoma of the pancreas (N = 50), ampulla (N = 19), distal bile duct (N = 10), and duodenum (N = 9). Forty-nine patients were men, 39 were women, and the mean age was 58 years (range: 34-84 years). The patients were divided into two groups on the basis of two different time periods: those operated on from 1969 to 1980 (N = 41) and those operated on from 1981 to 1986 (N = 47). There were no significant differences between the two groups in terms of mean age, sex distribution, duration of symptoms before presentation, or mean weight loss. Likewise, preoperative laboratory data were similar for both groups of patients. In addition, mean tumor size for patients with pancreatic cancer (3.5 cm vs. 3.2 cm) and patients with nonpancreatic periampullary cancer (1.9 cm vs. 2.2 cm) was similar in both groups, as was the incidence of positive lymph nodes. Among the 41 patients operated on during the first period, hospital morbidity and mortality rates were 59% and 24%, respectively. In contrast, hospital morbidity and mortality rates were 36% and 2%, respectively, among the 47 patients operated on during the recent period. During the recent period, more Whipple procedures were performed each year (7.8 vs. 3.4) and by fewer surgeons (3.4 operations/surgeon vs. 1.9 operations/surgeon). In addition, between 1981 and 1986, there were fewer total pancreatectomies (9% vs. 39%), fewer vagotomies (26% vs. 76%), and more pyloric-preserving procedures (30% vs. 0) performed compared with the earlier period. During the recent period, mean operative time (7.8 vs. 9.0 hours), mean estimated blood loss (1694 vs. 3271 mL), and mean intraoperative blood replacement (3.6 vs. 6.3 units) were all significantly less than in the earlier period. These findings suggest that the recent decline in operative morbidity and mortality may be due to fewer surgeons performing more Whipple resections in less time and with less blood loss. The actuarial 5-year survival rate for the 38 patients with nonpancreatic periampullary cancer was 34%. Surprisingly, the actuarial 5-year survival rate among the 50 patients with pancreatic cancer was 18%. Moreover, in the absence of positive lymph node involvement, the 5-year actuarial survival rate among patients with pancreatic cancer was 48%. No explanation is obvious for the improvement in survival among patients with pancreatic cancer.