Summary1. The initial response after endotoxin administration (3 mg/kg) in cats involved pulmonary vasoconstriction. This was not seen when endotoxin was given by slow infusion and it could be prevented after a bolus injection of endotoxin by pretreatment of the cats with aspirin (10 mg/kg). Intense mesenteric vasoconstriction occurred in all the cats. 2. The mesenteric vasoconstriction was a specific response of the mesenteric blood vessels. At the time the mesenteric bed constricted, the renal bed dilated, the hepatic arterial bed remained unchanged and the smooth muscle of the intestinal wall relaxed. 3. Arterial blood from cats with a fully developed mesenteric vasoconstriction after endotoxin administration was perfused through a normal intestine. No immediate vasoconstriction developed but the perfused intestine constricted slowly over 60 minutes. This suggests that mesenteric constriction was not due to circulating vasoconstrictor factors or the intestinal innervation, but involved a slow local mechanism within the intestine. It could not be prevented or reversed by a variety of pharmacological agents. 4. These observations suggest that endotoxin caused a unique type of mesenteric vasoconstriction in cats by a local mechanism which took up to 60 min to develop, was sufficiently potent to reduce mesenteric flow to <30% control, and was maintained until death of the cats. Blood from these animals did not clot when placed in a glass tube. 5. The mesenteric constriction and the clotting defect could be prevented by repeated administration of aminophylline and dextran solution before and after a bolus intravenous injection of endotoxin. Arterial pressure and mesenteric flow were maintained for at least 10 h in these experiments. Inadequate treatment intensified rather than reduced the intestinal mucosal damage. 6. Cats were treated with aspirin, endotoxin and the optimal regimen for prevention of the mesenteric constriction and allowed to recover from the anaesthetic agent. In this series, 63% survived indefinitely compared to 25% after aspirin and endotoxin treatment and 0% after endotoxin alone. 7. The possible mechanisms of action of aspirin and aminophylline-dextran solution are discussed. Our failure to obtain 100% survival is probably due to pulmonary damage which develops 10-24 h after endotoxin administration. This delayed pulmonary action of endotoxin is not prevented by aspirin M. M. Cohen and others treatment and it seenis unlikely that aspirin will be of any value in the treatment of the pulmonary lesion in man.