BACKGROUND AND STUDY AIMS: In previous randomized trials, early endoscopy improved the outcome in patients with bleeding peptic ulcer, though most of these studies defined "early" as endoscopy performed within 24 hours after admission. Using the length of hospital stay as the primary criterion for the clinical outcome, we compared the results of endoscopy done immediately after admission (early endoscopy in the emergency room, EEE) with endoscopy postponed to a time within the first 24 hours after hospitalization, but still during normal working hours ("delayed" endoscopy in the endoscopy unit, DEU). PATIENTS AND METHODS: We conducted a retrospective analysis of data from 81 consecutive patients with bleeding peptic ulcer admitted in 1997 and 1998 (age range 16 - 90 years). Of these 81 patients, 38 underwent DEU (the standard therapy at the hospital) and 43 underwent EEE. Patients in the two groups were comparable with regard to admission criteria, were equally distributed with respect to their risk of adverse outcome (assessed using the Baylor bleeding score and the Rockall score), and differed only in the treatment they received. Endoscopic hemostasis was performed whenever possible in all patients with Forrest types I, IIa, and IIb ulcer bleeding. RESULTS: We found similar rates in the two groups for recurrent bleeding (16 % in DEU patients vs. 14 % in EEE patients), persistent bleeding (8 % in DEU patients vs. none in EEE patients), medical complications (21 % in DEU patients vs. 26 % in EEE patients), the need for surgery (8 % in DEU patients vs. 9 % in EEE patients), and the length of hospital stay (5.1 days for DEU patients vs. 5.9 days for EEE patients). None of the differences between the two groups in these parameters were statistically significant. None of the patients died. CONCLUSIONS: Early endoscopy in an emergency room did not improve the clinical outcome in our 81 consecutive patients with bleeding peptic ulcer.