One hundred and thirty-two patients with endoscopically confirmed duodenal ulcers were entered into a 4-6-week double-blind trial of famotidine (Fm) and ranitidine (Rn). Seventy patients were randomised to Fm 40 mg at night and Rn placebo and 62 to Rn 300 mg at night and Fm placebo. Gelusil tablets were allowed for ulcer pain, and diary cards were issued. Clinical evaluations were carried out on entry and at 2 and 4 weeks, and endoscopy was repeated at 4 weeks. Patients with an unhealed ulcer at 4 weeks were continued on the same treatment for another 2 weeks and further endoscopy was carried out at 6 weeks. The groups were comparable with respect to age, sex, duration of the most recent attack, initial ulcer size and smoking, but the duration of disease tended to be longer in the Rn-treated group (P less than 0.05). Of the patients available for analysis, 75% of the Fm-treated and 78% of the Rn-treated patients were healed at 4 weeks and 91% of patients in both treatment groups were healed at 6 weeks. Gelusil consumption and symptomatic responses were comparable in the two groups, and no side-effects were noted apart from mild dizziness in 1 patient after a single Fm tablet. The mean initial size of ulcers healed at 4 weeks was significantly less than that of ulcers which were unhealed at 4 weeks (P less than 0.01), and the healing rate among smokers was significantly lower than that among non-smokers--48 of 72 patients (67%) and 41 of 45 (91%) respectively (P less than 0.01). The study shows that Fm 40 mg at night is as effective as Rn 300 mg in duodenal ulcer healing.