Purpose: To evaluate retrospectively the efficacy and chronic toxicities of concurrent radiotherapy and chronomodulated infusion 5-fluorouracil (5-FU) in patients with pancreatic adenocarcinoma. Methods and Materials: Twenty-eight patients with pancreatic adenocarcinoma were treated between January 1997 and May 2000 with 5-FU chronomodulated chemoradiotherapy. Chronomodulated delivery of chemotherapy was chosen on the basis of a lower toxicity profile in the treatment of GI malignancies. The median age was 64 years. Of the 28 patients, 12 were men and 16 were women. Eight patients had unresectable disease and 20 were treated after pancreatic resection. The median radiation dose was 50.4 Gy given in 28 fractions. The median field length and width was 10.6 cm and 10.9 cm, respectively. Concurrent chemotherapy with 5-FU was administered 5 d/wk, with a median total dose of 8.4 g/m2 (300 mg/m2/d). Chronomodulated 5-FU delivery consisted of a low basal infusion for 16 h followed by an 8-h escalating-deescalating infusion peaking at 10 pm. Survival and recurrence data were evaluated using Kaplan-Meier actuarial analysis. Toxicities were recorded using the Radiation Therapy Oncology Group grading system. Results: The median follow-up for all patients was 26 months (range, 4-68 months). The median overall survival for the 20 patients treated postoperatively was 34 months, with a 3- and 5-year actuarial survival rate of 40% and 21%, respectively. If the 3 patients with carcinoma of the ampulla were removed from the data set, the mean overall survival in the resected patients was 34 months, with a 3-year and 5-year actuarial survival rate of 40% and 17%, respectively. The 8 unresectable patients had a median overall survival of 14 months, and none lived past 2 years. No patient experienced Grade 3 or 4 hematologic toxicity or weight loss. Five patients had nausea and dehydration requiring i.v. fluids; only one (4%) was hospitalized. Four patients required a dose reduction of 5-FU, one for nausea, one for a transient ischemic attack, one for an infection, and one because of myocardial infarction. Seven resected patients, four of whom had no evidence of disease, developed diabetes mellitus 1-2 years after radiotherapy. Conclusion: Chronomodulated 5-FU administration, based on the concept of chronotolerance, has relatively low acute toxicity. Our median survival rate was greater than that after most chemoradiotherapy programs that result in more acute toxicity. Additional study is warranted to evaluate chronomodulated radiosensitizing chemotherapy schedules in prospective trials and with attention to late effects after radiotherapy, including diabetes mellitus. © 2005 Elsevier Inc.