Gastric haemorrhage was produced regularly in mice within 6 hours of the subcutaneous injection of a large dose (2 to 10 mg./kg.) of reserpine or of deserpidine. Rescinnamine,, and tetrabenazine (Ro 1-9569) were less active. Gastric haemorrhage was also produced within 6 hours when 5-hydroxytryptamine (10 mg./kg.) was injected every half-hour. Neither reserpine nor 5-hydroxytryptamine produced gastric haemorrbage in mice which had been vagotomized by tying the oesophagus at the cardio-oesophageal junction or which had been pre-treated with iproniazid. Amphetamine was less effective than iproniazid in preventing gastric haemorrhage after reserpine, and the following drugs were ineffective: cocaine, methyl phenidate (Ritalin), amarin, caffeine, nikethamide, lysergic acid diethylamide and its 2-bromo derivative (BOL148). Gastric haemorrhage was not observed in mice which had been given substantial doses of atropine or of hexamethonium before reserpine. The incidence of haemorrhage was substantially reduced by treatment with an antacid mixture. It is concluded that reserpine-like drugs cause gastric haemorrhage by a mechanism which has an important central component and which involves the liberation of 5-hydroxytryptamine.Haverback and have shown that gastric erosion was produced in rats within 18 hr. of the injection of large doses of 5-hydroxytryptophan. Reserpine in large doses had a similar effect (Benditt and Wong, 1957;Haverback and Bogdanski, 1957;Blackman, Campion, and Fastier, 1958). Some of the pharmacological actions of both these drugs appear to be mediated by 5-hydroxytryptamine. Reserpine liberates 5-hydroxytryptamine from platelets and certain other parts of the body, and interferes with the uptake of 5-hydroxytryptamine, while 5-hydroxytryptophan is the natural precursor to 5-hydroxytryptamine (Pletscher, Shore, and Brodie, 1955; Page, 1958). Therefore, when we observed in experiments on mice that large doses of reserpine produced gastric haemorrhage and melaena within a few hours (Blackman et al., 1958), we supposed at first that this was a peripheral effect brought about by an increase in " free " 5-hydroxytryptamine. The possibility that it was a central effect was raised by our observation that reserpine did not produce gastric haemorrhage in mice which had been vagotomized at the cardio-oesophageal junction. Another observation at variance with our original hypothesis was that reserpine did not produce gastric haemorrhage in mice which had been previously treated with iproniazid (Blackman et al., 1958). Since iproniazid is a powerful inhibitor of mono-amine oxidase and is believed to delay the destruction of 5-hydroxytryptamine in the body (Page, 1958), it would be expected to enhance rather than antagonize the harmful effect of reserpine on the gastric mucosa, were the erosion brought about by the local action of 5-hydroxytryptamine. These anomalous findings induced us to carry out the experiments reported here. METHODS Male mice weighing 30 to 40 g. were used. During the 24 hr. period precedi...