Both feedforward control by a signal descending from higher brain centres (termed central command) and feedback control by an afferent signal from contracting skeletal muscle (termed exercise pressor reflex) play a role in autonomic regulation of the cardiovascular system during exercise. Extensive investigations about exercise pressor reflex, conducted using anesthetized animals and decerebrate animals, revealed the characteristics of afferent and efferent activity and reflex centres in the brain stem. However, little is known regarding central command, particularly 1) the characteristics of the centrally-induced responses in sympathetic and parasympathetic outflows, 2) an interaction between central command and arterial baroreflex, and 3) the CNS origin(s) responsible for the central command. The reason is that it is difficult to conduct a comprehensive study during voluntary or spontaneous exercise in conscious or unanaesthetized animal preparations. Recently we have conducted a series of experiments to identify the above-mentioned questions using conscious, intact animals and unanaesthetized, decerebrate animals. All surgical procedures (the decerebration, implantation of an electrode on the renal nerve for measuring renal sympathetic outflow, and insertion of catheters into the carotid artery and vein) were conducted under halothane anaesthesia. We found that renal sympathetic outflow and heart rate abruptly increased immediately before or at the onset of spontaneous body movement or locomotion in decerebrate animals as well as conscious animals. The finding led us to a framework that central command may be generated in the caudal part of the diencephalon and/or regions within the brainstem, which may be triggered by descending output from the cortical structures, because the decerebration disconnects the cerebral cortex and the rostral diencephalon from the lower brainstem. Our recent study1 also provided a surprising result that a mechanical component of the exercise pressor reflex is suppressed in the conscious condition but becomes evident in the anaesthetized condition with pentobarbitone and propofol or the decerebrate condition, suggesting an influence of anaesthesia and cortical output on the exercise pressor reflex. When assessing the effect of central command on the baroreflex bradycardia in response to aortic nerve stimulation given at various phases before and during exercise, we revealed that central command interacts with the arterial baroreflex and produces a selective inhibition of the cardiac component of the baroreflex2, which may in turn evoke cardiac sympathetic stimulation for cardiac acceleration, rather than cardiac parasympathetic withdrawal3. Finally, regarding the CNS origin(s) responsible for the central command, we would like to propose a hypothesis that the mesencephalic ventral tegmental area (VTA) may be crucial for generating central command and linking the autonomic nervous and motor systems, because synchronised activation of renal sympathetic outflow and somatic motor bursts are produced by chemically stimulating neurons in the VTA4.