We have shown that a dipeptidyl peptidase 4 (DPP-4) inhibitor suppresses atrial remodeling in a canine atrial fibrillation (AF) model. Glucagon-like peptide-1 (GLP-1) is increased by DPP-4 inhibitors. However, it is not clear whether GLP-1 is involved in the suppression of atrial remodeling. In this study, we evaluated the effect of liraglutide (a GLP-1 analog) on atrial electrophysiological changes using the same canine AF model. We established a canine AF model using continuous 3-week rapid atrial stimulation in seven beagle dogs divided into two groups: a liraglutide group with four dogs (3-week atrial pacing with liraglutide (150 µg/kg/day) administration) and a pacing control group with three dogs (3-week pacing without any medicine). We evaluated the atrial effective refractory period (AERP), conduction velocity (CV), and AF inducibility every week during the protocol using implanted epicardial wires against the surfaces of both atria. In the pacing control group, the AERP was gradually shortened and the CV was decreased along the time course. In the liraglutide group, the AERP was similarly shortened as in the pacing control group (94 ± 4% versus 85 ± 2%, respectively; p = 0.5926), but the CV became significantly higher than that in the pacing control group after 2 and 3 weeks (95 ± 4 versus 83 ± 5%, respectively; p = 0.0339). The AF inducibility was gradually increased in the pacing control group, but it was suppressed in the liraglutide group (5 ± 9% versus 73 ± 5%; p = 0.0262). Liraglutide suppressed electrophysiological changes such as AF inducibility and CV decrease in our canine AF model.