1. Ventricular mechanoreceptors are known to exist and can when stimulated induce reflex vasodilatation, but the nature of the effective stimuli and the physiological role of the reflex remain to be established. 2. Dogs were anaesthetised with chloralose and a cardiopulmonary bypass established. Ventricular pressures were separated from those in the aortic root and coronary arteries by a balloon inflated in the ventricular outflow tract. Ventricular filling was controlled by adjusting the rate of inflow of blood through an apical cannula and peak pressure by regulating the outflow pressure from the same cannula. Carotid and aortic pressures were also controlled and vascular resistance was assessed from changes in perfusion pressure (constant flow conditions) to the descending abdominal aorta. 3. Increased coronary or carotid sinus pressure induced a significant vasodilatation. Changes in ventricular peak systolic pressure, without associated changes in end-diastolic pressure, had no significant effect on vascular resistance. In contrast, changes in end-diastolic pressure did induce vasodilatation that, although small, was proportional to the magnitude of the end-diastolic pressure change. 4. Changes in ventricular inotropic state induced by dobutamine infusion or by stimulation of efferent cardiac sympathetic nerves did not induce significant responses. Furthermore, the combined effects of reduced ventricular filling and increased inotropic state were also ineffective in inducing responses. 5. We conclude that, to induce reflex responses, the only effective stimulus to ventricular mechanoreceptors was an increase in filling. Compared with other mechanoreflexes, however, responses to ventricular distension were small and seem unlikely to be of importance except perhaps during abnormal ventricular distension.