The existence of abdominal arterial baroreceptors has long been controversial. Previously difficulties have been encountered in localizing a stimulus to abdominal arteries without affecting reflexogenic areas elsewhere. In these experiments, using anaesthetized dogs, the abdomen was vascularly isolated at the level of the diaphragm, perfused through the aorta, and drained from the inferior vena cava to a reservoir. Changes in abdominal arterial pressure were effected by changing the perfusion pump speed. During this procedure the flow back to the animal from the venous outflow reservoir was held constant. Increases and decreases in abdominal arterial pressure resulted, respectively, in decreases and increases in perfusion pressure to a vascularly isolated hind-limb and in some dogs also a forelimb. Responses were significantly larger when carotid sinus pressure was high (120-180 mmHg) than when it was low (60 mmHg). Responses were still obtained after cutting vagus, phrenic and splanchnic nerves, but were abolished by spinal cord lesion at T12. These experiments provide evidence for the existence of abdominal arterial baroreceptors. The afferent pathway for the reflex vasodilatation appears to run in the spinal cord.