This study was designed to determine (1) whether the site of biopsy within the proximal small bowel affects the ability to assess mucosal architecture in general, or to confirm a diagnosis of celiac sprue specifically; and (2) whether endoscopic small bowel biopsy using standard forceps can obtain adequate biopsy specimens to detect or exclude mucosal abnormalities. Three-hundred fifty-two biopsy specimens were obtained prospectively from 26 patients (8 sprue, 2 nonspecific changes, 16 normal) with "jumbo" and standard forceps from jejunum, ligament of Treitz, fourth, third, and second portions of the duodenum. There was no difference in biopsy specimen quality from different locations. All celiac sprue patients had at least one good or excellent specimen from each location, thereby allowing the diagnosis to be made equally well from second, third, and fourth portions of the duodenum, as well as at the ligament of Treitz and jejunum. No false-positive diagnoses of celiac sprue were made. Finally, the standard biopsy forceps provided good or excellent specimens in all patients.