Background: The time-trend in the survival of elderly pancreatic cancer patients was still unclear. Thus, the aim of this study was to compare the survival benefit of young and elderly pancreatic cancer patients by a time-trend analysis. Methods: From 2004-2013, we obtained 5,341 of young patients (< 80 years) and 569 elderly patients (≥ 80 years) from the Surveillance, Epidemiology, and End Results (SEER) database, and the overall survival of these patients were analyzed by Kaplan-Meier estimator. The independent factors which could predict the survival of patients were determined by cox proportional hazards model.Results: We observed that the median overall survival of the young patients in 2004-2008 cohort was significantly (P < 0.001) increased when compared to that in the 2009-2013 cohort. However, we did not observe the survival benefit for the elderly patients. The Cox proportional hazards model demonstrated that the tumor size, lymph node ratio, grade, and AJCC TNM stage were independent factors of survival. Conclusions: This study demonstrated that compared to 2004-2008, the survival of elderly patients in 2009-2013 was not significantly improved. Thus, the clinicians still need to administer more care to elderly patients.