Clinical, endoscopic, and laboratory data were collected prospectively in 701 patients with bleeding peptic ulcer. The overall rebleeding rate was 16.1% and increased the risk of a fatal outcome by 17 fold (1.2% versus 20.6%, p less than 0.001). Rebleeding was documented in more than 75% of the group who did not survive following initial conservative management. Rebleeding was more likely (24.1% versus 14.2%, p less than 0.02) when shock was present on admission and the risk of a rebleed was doubled in patients over 60 years of age (22.1% versus 10.9%, p less than 0.001). Ulcers greater than 1 cm in size carried twice the risk of rebleeding (23.9% versus 12.4%, p less than 0.002). Concomitant medical illness had a significant adverse effect on outcome (p less than 0.05). Shock on admission was associated with a doubling of mortality figures (9.5% versus 3.7%, p less than 0.01). The identification of endoscopic stigmata of recent hemorrhage (ESRH) tripled the risk of mortality (7.5% versus 2.4%, p less than 0.002), ESRH were more frequently encountered when ulcer size was larger than 1 cm (61.4% versus 39.8%, p less than 0.001). Respective mortality rates for ulcers less than or equal to 1 cm and greater than 1 cm in size were 1.6% and 12.5% (p less than 0.001), corresponding mortality figures for patients over 60 years of age being 4.4% and 16.4% (p less than 0.002). The risk of a rebleed tripled (6.7% versus 2.6%, p less than 0.02) when ESRH were evident. There was a 6-fold increase in mortality following emergency surgery when compared with conservative management of patients in whom no surgical intervention was necessary (2.6% versus 14.9%, p less than 0.001). In summary, age over 60 years, previous medical illness, shock on admission, large ulcer size, and ESRH were each associated with an increased risk of rebleeding and mortality.