A set of evidence-based consensus guidelines for cardiopulmonary resuscitation (CPR) in dogs and cats (RECOVER guidelines) was published in 2012. The purpose of this study was to investigate the clinical outcomes of CPR performed according to those guidelines in dogs. A total of 141 dogs with cardiopulmonary arrest (CPA) were identified and underwent CPR between January 2012 and December 2015 at the Sapporo Nighttime Animal Hospital. CPR was performed according to no-consensus traditional veterinary CPR procedures in 68 dogs (TRADITIONAL group), and according to the RECOVER guidelines in 73 dogs (RECOVER group). There was no significant difference in the age, body weight, or time from CPA identification to initiation of CPR between the TRADITIONAL and RECOVER groups (median [range]: 10 [0-16] vs. 11 [0-16] years; 6.6 [1.0-58.6] vs. 5.5 [1.1-30.4] kg; and 0 [0-30] vs. 0 [0-30] min, respectively). In the TRADITIONAL group, 12 dogs (17%) achieved a return of spontaneous circulation (ROSC), but none survived to hospital discharge. However, 32 dogs (43%) in the RECOVER group achieved ROSC, and 4 dogs (5%) were discharged from the hospital. Incorporating the RECOVER guidelines into clinical practice significantly improved the ROSC rate (P<0.001). However, the rate of survival to hospital discharge was still low. This may suggest that a superior intensive care unit that provides advanced post-CPA care could benefit veterinary CPR patients.