Thirty-one patients with endoscopically proven chronic gastric ulcer completed a randomised double-blind trial comparing the effects of cimetidine and placebo on ulcer healing. Seventeen patients received cimetidine 400 mg bid and 14 patients received placebo. Repeat endoscopy at six weeks showed that the ulcer had healed in 12 patients (71%) receiving cimetidine and in four patients (29%) receiving placebo (p=O.032). Non-smokers healed their ulcers better than smokers (83% vs 35%, p=O.023). The use of cimetidine was not associated with any adverse effects.