Objective: To establish a fluoroscopic percutaneous vertebral augmentation model in dogs through measurement and analysis of canine spinal anatomy. Furthermore, the effectiveness and safety of this modeling method are assessed through post-operative radiological analysis. Methods: Morphological measurements and their parameters of the lumbar vertebrae of six dogs aged around 12 months were taken: height of the L1-L7 vertebrae, width of the vertebral base, distance from the upper edge of the intervertebral disc to the narrowest part of the vertebra, distance from the vertical line of the spinous process to the upper edge of the intervertebral disc, and the vertical distance from the midpoint of the transverse process to the lower edge of the intervertebral disc. This was done to clarify the anatomical characteristics of the canine vertebrae and to determine the optimal location, direction, and depth for bone cement injection. Subsequently, a percutaneous vertebral augmentation model was established in the L4, L5, and L6 vertebrae of six healthy Beagle dogs weighing between 20-25kg. Four weeks post-surgery, the subjects were euthanized and radiologically examined. Primary observations included: surgical duration, post-operative distribution of the implanted bone cement, and the integrity of the vertebral canal and anterior edge of the vertebrae. Results: Anatomical observation of the canine vertebrae revealed that the vertebral height gradually increased from L1 to L5 and then decreased from L5 to L7. The width of the vertebral base consistently increased from L1 to L7. The distance from the vertical line of the spinous process to the upper edge of the intervertebral disc showed an increasing trend from L1 to L7 (1.9 to 4.0 mm). The distance between the midpoint of the base of the transverse process and the lower edge of the intervertebral disc gradually increased from L1 to L5 (4.7 to 6.9 mm). There was no significant statistical difference in the distance between the midpoint of the base of the transverse process and the lower edge of the intervertebral disc in the L4, L5, and L6 segments among the dogs (P=0.925). The midpoint of the root of the transverse process of the spine was taken as the puncture point, and the insertion direction and horizontal plane were at an Angle of 20° - 30 °, with a head tilt of 5° - 15° and a puncture depth of 1.2 cm to 1.5 cm. If the puncture was directed towards the caudal side of the vertebra, the angle of the needle tail was 30° to 35°, with a penetration depth of 1.5 cm to 1.8 cm. This technique allowed for the successful construction of a canine vertebral puncture surgical model. A total of 15 canine vertebral puncture surgical models were successfully created, with an average surgery time of 22.7 ± 4.6 minutes (ranging from 15 to 30 minutes) per vertebral segment. During surgery, one vertebral segment experienced spinal cord injury resulting in paralysis of the hind limbs and incontinence of both bowel and bladder. Two vertebral cortical bones fractured, but there were no deaths due to anesthesia or infection. Four weeks post-surgery, Micro-CT three-dimensional reconstructions consistently showed bone cement distributed within the trabecular bone of the canine vertebrae, with newly formed bone tissue enveloping the implanted material. There was no leakage, and no complications such as damage to the vertebral canal or the anterior wall of the vertebrae were observed. Conclusions: Based on the anatomy of the canine lumbar vertebrae (L4 to L6), using the midpoint of the base of the transverse process as a bony landmark, a safe and reliable canine vertebral augmentation puncture model can be successfully established.