Extraperitoneal air insufflation done through a single presacral puncture is a simple procedure which is considered safe by those workers who have used it in large numbers of cases. Mosca (2) collected over 1,500 cases without fatal accidents. Ruiz Rivas (4), one of the originators of the method, stresses the fact that a needle inserted in the mid-line between the rectum and the sacrum encounters no large vascular channels. For this reason, gas embolism is considered unlikely. That it may occur is shown by the following case. A woman, 74 years of age, was subjected to extraperitoneal insufflation to localize a high retroperitoneal tumor. The needle was inserted in the usual manner, aspiration was done without obtaining blood, and the insufflation was started with oxygen at a pressure of 30 cm. of water. The patient was in the left decubitus position. In about fifteen minutes, 500 c.c. of oxygen were injected without difficulty. The patient was then placed in the prone position to obtain a better distribution of the gas. After the injection of another 100 c.c, of oxygen, she suddenly became dyspneic, then cyanotic. The skin was clammy, the blood pressure was unobtainable, and the pulse irregular and thread-like at 120 beats per minute. The insufflation was discontinued. Two minims of adrenalin were administered, and, in order that the heart beats might be followed, the patient was rolled over on her back. A loud “millwheel” sound could be heard with each beat, unmistakable evidence of gas in the heart. The patient had become very restless and anxious and resented the oxygen mask. Her distress was especially marked in the supine position. At this point, she was again turned on her left side. In a few minutes the “millwheel” sound disappeared and she became slightly more comfortable. The pulse rate was now 160. Cyanosis and apprehension continued for approximately thirty minutes. Gradually the pulse rate dropped to 104 per minute, the color became normal, and the patient relaxed. At no time was she unconscious. In fact, she recalled vividly the “suffocating sensations,” which were extreme in the prone and supine positions. About three hours later a film of the abdomen revealed no gas in the extraperitoneal spaces. The absence of gas in the extraperitoneal spaces after three hours is significant. Since oxygen is not reabsorbed in so short a time, one must postulate that the majority, if not all, of the 600 c.c. was injected directly into the venous system. The survival of the patient in spite of the intravenous injection of such a quantity of gas may be explained by the experimental work of Durant, Long, and Oppenheimer (1). These authors, working with dogs, found that with a slow rate of injection the animals could tolerate large amounts of intravenous gas. When gas was injected fairly rapidly, fatal embolism developed when the dogs were kept on the right side, on the back, or in the prone position.