Abstract: Intravenous injection of the aqueous extract of Salsola collina produced marked and prolonged hypotensive effect in anaesthetized dogs. Repeated administrations showed no marked tachyphylaxis. The drug had no influence on the hypotensive action of acetylcholine and the depressor response induced by electric stimulation of the peripheral end of vagus nerve. Its hypotensive effect was not blocked by previous administration of atropine, but reduced to some extent in vagotomized animals. Hence it had no effect on the M-cholinergic system. It did not reveal any adrenolytic and ganglion-blocking action. During the period of hypotension, the pressor responses induced by serial occlusions of common carotid arteries and electric stimulation of the central end of sciatic nerve were inhibited or eliminated. The hypotensive effect was not influenced by bilateral section of the sinus nerves. Besides, the drug gave a reflex vasodilating effect in perfused isolated rabbit's ear with the innervation intact. Therefore it is suggested that the hypotensive mechanism was mainly due to an inhibition of the vasomotor centre or the central sympathetic structure. The aquaeous extract of Salsola collina exhibited a pronounced sedative action. It reduced the spontaneous activity, prolonged the pentobarbital induced sleeping time and potentiated the effect of nonsleeping doses of chloral hydrate to induce sleeping, but had no effect on the convulsion and death caused by strychnine and metrazol. In mice, by the method of motor defense conditioned reflex, subcutaneous injection of Salsola collina 3g/kg had no marked influence on the conditioned reflexes. When the doses were raised to 5, 10g/kg, it prolonged the latency of conditioned reflexes, partly abolished the conditioned reflexes, but did not alter the differentiation. At the dosage of 20g/kg, this effect was more pronounced. In subcutaneous doses of 5, 10g/kg, it accelerated the disappearance of positive conditioned reactions. In mice the acute LD₅₀ by the subcutaneous route was found to be 56g/kg.