Volumes used in lumbosacral epidural injections for anesthesia have remained unchanged since the 1960s. The goals of this cross-sectional observational study were to characterize the three-dimensional spread of a lumbosacral epidural injection, as well as confirm that the commonly used volume of 0.2 ml/kg injected into the lumbosacral epidural space reaches the thoracolumbar (TL) junction in the majority (≥80%) of dogs. Ten clinically normal, adult, nonpregnant, mixed-breed dogs were obtained within five minutes of euthanasia and 0.2 ml/kg of radiopaque contrast medium was injected into the lumbosacral epidural space. A computed tomography scan of the TL spine was performed immediately following the injection. Migration of contrast reached the TL junction in 8 of 10 (80%) dogs. Contrast was well visualized in all epidural planes with contrast travelling predominantly in the dorsal epidural space in 7 of 10 (70%) dogs. There was no significant difference in the weight of dogs where the epidural injectate reached the TL junction and those where it did not (P = 0.16), or in the weight of dogs where the cranial-most point of the contrast column was in the dorsal versus the ventral epidural space (P = 0.32). This preliminary study supports the use of computed tomography to characterize injectate distribution in the canine thoracolumbar epidural space and provides evidence that a 0.2-ml/kg volume is likely to reache the TL junction in most dogs. Further studies are needed in live dogs to determine if variables affecting human epidural injectate doses have similar effects in the dog.