Understanding central circuits that control the cardiovascular system and how these circuits change with disease or injury is a major challenge. Sympathetic preganglionic neurons (SPN) in the intermediolateral cell column (IML) of thoracic and upper lumbar cord are critical neurons for blood pressure control because they provide central drive that regulates blood vessel diameter. SPN are topographically organized. Rostrally, they regulate targets in the upper body, such as the heart, whereas caudally, they control abdominal and pelvic viscera. Spinal cord injuries disrupt the connections between SPN and neurons above the lesion. The rostrocaudal topography of SPN means that injury location determines its autonomic sequelae. If damage occurs in the upper thoracic or cervical cord, blood pressure control can be profoundly disturbed. People and animals with such injuries experience hypotension and autonomic dysreflexia, a condition characterized by hypertensive episodes that are triggered by noxious or innocuous sensory input entering the cord below injury level. Dysreflexia is thought to occur because of loss of baroreflex input to SPN controlling the splanchnic vasculature, an important bed for reflex control of arterial pressure. My collaborators and I have revealed some of the major effects of spinal cord injury on SPN and their synaptic inputs. After a complete transection (under anaesthesia) at segments T4/5, SPN retract and then regrow their dendrites. Their cell bodies shrink and return to normal size. These changes in SPN correlate with the time required for clearance from the IML of axons severed by the transection and are completed by two weeks after injury. Significant changes to the synaptic input of mid-thoracic SPN also occur acutely. We compared the density of synaptic input to choline acetyltransferase-immunoreactive SPN in T8 from rats with intact cords and rats with 3- or 14-day transections and determined the amino acid content of inputs using immunogold labelling. At 3 days after injury, the number of synapses/10 µm of SPN membrane had decreased by 34% on somata but increased by 66% on dendrites. Almost half the inputs lacked amino acids. By 14 days, the density of inputs to dendrites and somata decreased by 50% and 70%, respectively. The proportion of input that contained glutamate was less in rats with 14-day injuries than in rats with intact cords whereas the proportion of input that contained GABA increased. Thus, in acute injuries, SPN participate in vasomotor control despite profound denervation. Furthermore, an altered balance of excitatory and inhibitory input may explain injury-induced hypotension. Innervation of the IML by other populations of neurochemically identified axons also changes caudal to a complete transection. In intact cord, supraspinal axons containing tyrosine hydroxylase (TH) or phenylethanolamine-N-methytransferase densely supply the IML and synapse on SPN. At 14 days post-transection, the density of these axons is substantially decreased although some TH synapses persist on SPN. By 11 weeks after injury, all of the catecholamine axons have disappeared from the IML. Serotonergic axons disappear more quickly. These supraspinal axons abundantly supply the IML of intact cord but are absent by 2 weeks after injury. Caudal to both acute and chronic (10-12 week) transections, axons containing substance P, enkephalin and neuropeptide Y are present in the IML. Synapses containing each of these neuropeptides occur on SPN from acutely transected cord. These observations indicate that, in injured cord, the IML and SPN receive synaptic input from neuropeptide-containing spinal interneurons. Interestingly, the anatomical consequences of spinal cord injury appear to be quite different for caudal SPN retrogradely labelled from the major pelvic ganglion, which contains post-ganglionic neurons innervating the bladder, lower bowel and reproductive organs. Enkephalin- or galanin-containing axons each supply more than half of the innervation of these SPN and a transection does not appear to produce a significant change in their proportions. Thus, more rostrally located SPN are predominantly controlled by supraspinal neurons whereas SPN that control the pelvic viscera are mainly regulated by spinal interneurons. The dominance of intraspinal control for pelvic visceral SPN has important implications for people with spinal cord injuries. Drugs that target persisting interneuronal pathways may be effective treatments for sympathetically-mediated pelvic dysfunction. Moreover, since circuitry controlling pelvic visceral SPN is not significantly affect by injury, restoring sympathetic control of pelvic organs may be less difficult because re-establishment of direct synaptic input from regrowing supraspinal axons may not be so critical.