Objective: To evaluate the accuracy of ultrasonographic guidance for injection near the ventral rami of the cervical spinal nerves (VRCSN).Study design: Ex vivo study. Sample population: Ten cadavers (n = 40 injections).Methods: Left and right VRCSN at C4/5 and C5/6 were infiltrated at the intervertebral foramen (IVF) under ultrasonographic guidance. A vertically orientated 3.5-MHz curvilinear probe was used to identify the craniocaudal midpoint of the articular process joint (APJ). The probe was moved ventrally until the IVF was identified. An 18-gauge 15-cm-long spinal needle was aimed immediately below the ventral margin of the cranial articular process, where 1 mL of latex was injected. Cadavers were dissected, and the location of the latex was recorded relative to the ventral ramus, the spinal cord, and the vertebral artery.Results: Direct infiltration of the ventral ramus occurred in 73% (29/40) of injections; 15% (6/40) of injections were < 0.5 cm from the ventral ramus, and 13% (5/40) of injections were > 0.5 cm (maximum 1.5 cm). Latex was located at a mean of 2.4 cm (range, 0.8-4) from the spinal cord and 0.9 cm (range, 0-1.5) from the vertebral artery. Conclusion:Ultrasonographic guidance of perineural injections of the caudal VRCSN was repeatable and accurate, with 88% of injections located within 0.5 cm of the ventral ramus. Iatrogenic damage was limited to one penetration into the vertebral artery.Clinical significance: Ultrasonographic guidance of perineural injections of the caudal VRCSN may be considered by operators experienced in cervical APJ scanning in horses with suspected cervical spinal nerve compression.