OBJECTIVE To examine the anatomy of the lumbar epaxial region and to describe two different ultrasound-guided approaches for the lumbar erector spinae plane (ESP) block in dogs. METHODS An anatomical and experimental cadaver study. METHODS A group of 19 canine cadavers. METHODS The anatomy was described following dissection of two cadavers. Bilateral ultrasound-guided ESP injections with 0.4 mL kg-1 of contrast dye were performed in 17 adult Beagle cadavers using either transversal (TVS) or parasagittal (PST) approaches. Computed tomography was performed to measure the total length of the contrast dye column and the epidural, intravascular, hypaxial and intra-abdominal migration. Dissections were performed to assess the spread of the contrast dye and to determine the degree of staining of the dorsal branches of the spinal nerves (DBSN). Mann-Whitney U and chi-square tests were used to compare data between groups. RESULTS Using both techniques, the contrast dye was observed within the ESP compartment. There was no difference in the total length of the contrast dye column between TVS and PST approaches (p = 0.056). Using the TVS approach, multisegmental staining of the DBSN was visible with 100% (17/17) of injections, while complete staining of the DBSN was achieved at 94% of the injection sites. Using the PST approach, these values were 29% (5/17) and 23% (4/17), respectively. The TVS approach stained more DBSN than the PST approach (p = 0.001), with a median (range) of 2 (2-3) versus 0 (0-3) DBSN, respectively. Using the TVS approach, epidural and intravascular migration were present in 2/17 (p = 0.485) and 3/17 (p = 0.227) injections, respectively. CONCLUSIONS Both ultrasound-guided approaches resulted in a spread of the contrast dye within the ESP compartment. Although there were no differences in the total length of the contrast dye column, the TVS approach was superior to the PST approach in staining DBSN.