The objective of this study was to determine the effect, if any, on survival of adjuvant 5-FU-based chemoradiotherapy following pancreaticoduodenectomy for pancreatic carcinoma. A systematic review of the published literature was undertaken. Survival estimates were derived from published reports. Five prospective studies (4 level I, 1 level II) with a total of 607 (229 surgery only; 378 surgery-adjuvant) patients followed for survival met selection criteria. Two-year survival ranged from 15%-37% in the surgery only group and 37%-43% in the surgery and adjuvant groups. The survival advantage (absolute difference) ranged from 3%-27% and no individual study achieved statistical significance (5%). Although clinical heterogeneity existed in surgery-alone control groups with regard to trial date, no statistical heterogeneity was detected (P = 0.459, chi2 test), allowing pooling of survival data. Using a fixed effects model, the summary estimate showed an absolute 2-year survival benefit with adjuvant therapy of 12% (95% CI, 3%-21%, P = 0.011). Trials after 1997 (n = 3) indicated a survival benefit of 8% to patients receiving adjuvant therapy (95% CI, -3-18%, P = 0.145). The result was not statistically significant, and there was no evidence of heterogeneity (P = 0.626, chi2 test). Summary estimates were unchanged when the analysis was performed with a random effects model. 5-FU based chemotherapy with radiotherapy given after resection imparts a small overall survival benefit of 2 years. The benefit of 5-FU-based adjuvant therapy, however, has declined in recent years, and its significance remains unproven in the context of current diagnostic and surgical practice.