104 consecutive patients with upper gastrointestinal bleeding admitted to a district general hospital over a period of 20 months were considered for laser photocoagulation. 58 patients were found to have peptic ulcers, of which 42 had stigmata of recent haemorrhage accessible to laser photocoagulation. 17 were treated with a Nd-YAG laser and 25 were managed conservatively. There was no significant difference in the number of patients who re-bled and required surgery, mortality rate, length of stay in hospital or the amount of blood transfusion required in either group. We suggest that the routine use of a laser for bleeding peptic ulcers in a district general hospital is not justified.