A retrospective study has been made of the diagnosis and treatment of 250 patients with carcinoma of the pancreas (confirmed by celiotomy, biopsy, and autopsy in 242) from a private surgical group at one hospital (Series A), a 12-year medical-surgical experience at a second private hospital (Series B), and a 10-year experience at the University of Nebraska Hospital, which included a high percentage of low-income referrals (Series C). Diagnosis was usually delayed until obstructive jaundice or radiologic evidence of upper gastrointestinal tract deformity existed, and was usually achieved by celiotomy. Follow-up studies using second look operations and autopsies suggest that supervoltage radiation therapy to at least 3,000 rads tissue dose initial course combined with initial 5-fluorouracil (FU), followed by weekly FU maintenance therapy, has at times favorably modified the course of the primary as well as locally metastatic carcinoma. Eight of nine patients whose erythrocytes were tagged with Cr51 for quantitative specific blood loss measurement excreted one or more stools significantly exceeding the normal 0.6 ± 0.4 ml lost per 24 hours; the ninth patient had complete biliary obstruction at the time of testing. In two patients, the blood loss was as high as 83-239 ml daily due to extrinsic gastric invasion. In a third patient, the blood loss decreased from 50 ml daily to 4 ml daily during irradiation and FU therapy. Survivorship studies indicate no significant difference between Series A and C cases in this study, with the proportion of survivors at 6 and 23 months palliated by major surgical resection attempts beginning to be equalled at 23 months, by those who received only bypass operations combined with pancreatic irradiation to 3,000–4,000 rads and FU therapy. Patients over 45 with nonspecific digestive disturbances and negative endoscopic and radiologic examination with persistent daily gut blood loss by the radiochromium technique of 1.8 ml or greater for 2–3 weeks duration should have the level of bleeding identified either by sequential gastrointestinal intubation and serial aspiration of digestive secretions for blood loss measurement, or by radiologic methods, which are usually negative early in the growth of extra-ampullary carcinomas.