SUMMARY1. Stress activates the hypothalamus causing central adrenergic discharge and stimulation of the autonomic sympathetic system. Reserpine produces the same effect and, therefore, its acute gastric mucosal injury is stress-induced. This injury was employed in the gastric diversion rat, a model for determining gastric acid secretion under basal conditions, to examine the relationship of the vagus nerve to the autonomic sympathetic system in the mechanism of stress-induced acute gastric mucosal injury.2. After 6 h of reserpine (5 mg/kg i.P.), all rats developed oval or round lesions confined to the glandular stomach and of no constant relationship to rugal crests (lesion score 29+ 2-7 mm2, mean +S.E., n = 10). Microscopically, these lesions were vascular in origin, developing as intramural foci of haemorrhage or necrosis and expanding to communicate with the lumen. Pre-treatment with potent antisecretory doses of the anticholinergic atropine (5 mg/kg i.P.) or the H2-receptor antagonist cimetidine (40 mg/kg i.P.) did not influence this reserpine action (28 + 3 mm2 and 27-5 + 2'3 mm2, respectively, mean+ s.E., n = 10). Protection against the reserpine lesions by the a-adrenoceptor blocking drugs phenoxybenzamine or phentolamine given in a dose of 10 mg/kg i.P. was significantly (P < 0 01) more than that afforded by the 5 mg/kg i.P. dose. However, the 15 mg/kg i.P. dose was completely protective against the lesions. Vagotomy had a similar protective effect. Interruption of autonomic sympathetic delivery to the stomach by coeliac ganglionectomy had no influence on the macroscopic or microscopic effects of reserpine on the stomach (30 5 + 3-4 mm2, mean +s.E, n = 10).3. The H+ output associated with 6 h of gastric diversion (61 + 4-5 ,tmol, mean +S.E.) was significantly (P < 0-001) depressed by reserpine alone (26 + 2 jtmol) or with atropine (19+1-8,imol) or cimetidine (21+2,tmol). Protection against the reserpine lesions by phenoxybenzamine or phentolamine was associated with dosedependent increase of H+ output, which with the 15 mg/kg dose was similar to that of control values (58 + 4-1 ,tmol and 60-3 + 2-8 ,mol vs. 61 + 4-5 /tmol). Vagotomy protection was associated with an H+ output significantly (P < 0-001) lower than that with reserpine alone (14 + 1-4 jumol). Coeliac ganglionectomy had no influence on the H+ output associated with reserpine treatment.4. The results of this study suggest that in the rat stress produces vagal delivery of a-adrenoceptor stimulation to the stomach, causing vasoconstriction which depresses H+ secretion and induces mucosal injury. The autonomic sympathetic system does not participate in this stress-induced action. Since the gastric diversion rat was used in this study, the results also show that duodenal contents are not required for the development of stress-induced gastric mucosal injury and suggest that H+ is not a critical factor in its mechanism.