1. In cats anaesthetized with pentobarbitone sodium, atropinized by i.v. atropine methyl nitrate and artificially ventilated, experiments were carried out (a) to localize the site where glycine acts on the ventral surface of the medulla when, on topical application through paired Perspex rings caudal to the trapezoid bodies, it produces a fall in arterial blood pressure, (b) to compare the effects of uni- and bilateral application, and (c) to study the blood pressure effects produced by electrolytic lesions of the glycine-sensitive areas.2. Blood pressure fell only a little on unilateral application of glycine through one of the Perspex rings, but a pronounced fall occurred on its bilateral application. The fall was too large to be explained by two minimal responses added together. Thus the application of glycine to one side potentiated the depressor effect of glycine applied to the other side.3. By moving the paired Perspex rings rostrally or caudally to different positions on the medulla and determining for each position the effectiveness of glycine, the glycine-sensitive areas were found to be restricted to a 1.5 mm wide strip situated 1-2.5 mm caudal to the trapezoid bodies. By making electrolytic lesions within the limits of this strip the glycine-sensitive areas were found to be not wider than 1.5 mm in the mediolateral direction and to be situated about 4 mm lateral to the mid line. Histologically, such lesions involved the cells of the parvicellular part of the lateral reticular nucleus.4. Placing an electrode, with a diameter of 1 mm, under light pressure on the glycine-sensitive area produced a short-lasting steep rise in blood pressure. The same effect was produced when a current was passed through the electrode to destroy the underlying tissue, but after its destruction the passage of current no longer produced the pressor effect.5. Once the glycine-sensitive area of one side was destroyed, glycine applied to the destroyed area through one of the Perspex rings no longer produced its small depressor effect, but when applied to the intact area of the other side, a pronounced fall in blood pressure occurred. Thus unilateral destruction had the same effect as unilateral glycine application. It potentiated the depressor effect of glycine applied to the other side.6. Following unilateral destruction of the glycine-sensitive area there was only a small fall in blood pressure; following its bilateral destruction blood pressure fell to a low level.7. It is concluded that the cells in the small circumscribed glycine-sensitive areas on the ventral surface of the medulla may play a key role in the maintenance of arterial blood pressure and that the cells of one side are sufficient for this function.