Objective: To use a population-based dataset to evaluate length of stay and the number of and reasons for readmission in Medicare patients undergoing pancreatectomy. Methods: We used SEER-Medicare linked data to evaluate the initial hospitalization, readmission rates within 30 days of and up to one year after initial discharge, and reasons for readmission in patients 66 years and older undergoing pancreatectomy. After excluding deaths during the initial hospitalization, patients requiring readmission within 30 days were compared to those who did not.Results: 1,730 Medicare patients underwent pancreatectomy for pancreatic cancer between 1992 and 2003. There were 130 in-hospital deaths (7.5%). Of the 1600 patients who survived until discharge, 784 patients were readmitted a total of 1,766 times within one year of discharge. The length of stay and readmission rates remained constant over the time period. Within 30 days of discharge, 248 patients (15.5%) were readmitted a total of 320 times. Early readmissions were primarily related to postoperative complications. The most common reasons for readmission were abscess/sepsis in 20%, delayed gastric emptying in 15%, chemotherapy/recurrence related in 10%, dehydration in 9%, pancreatic fistula/bile leak in 8%, cardiovascular complications in 5%, hemorrhage/GI bleed in 3%, and wound infection in 2%. 10% of readmissions were unrelated to surgery. Readmissions after 30 days were increasingly related to progression of disease. Patient age, gender, race, comorbidities and tumor stage did not predict early readmission within 30 days. The 30-day mortality rate was 15.6% in patients requiring readmission and 14.9% in patients who did not (P = 0.75). However, patients readmitted within 30 days of discharge had worse long-term survival (median = 11.8 months, 5-year survival 18%) than patients not requiring readmission (median = 16.5 months, 5-year survival 18%, P = 0.04). Readmission within one year was associated with a much worse median survival (21.6 months vs. 12.1 months, P<0.0001) related primarily to progression of disease requiring readmission. Conclusions: Our study demonstrates a one year readmission rate of 49% and a 30-day readmission rate of 15% after pancreatectomy. Previous studies do not report 30-day readmission rates, which are most important as these rates reflect surgical complications. These data demonstrate that readmitted patients are not more likely to die within 30 days. However, their survival is decreased in the initial year following surgery. This is likely attributed to the surgical complications which required readmission.