The sinoatrial node (SAN) is the heart’s pacemaker. The human SAN anatomy is complex. There is accumulating evidence that the mammelian cardiac anatomy consists of an electrical insulating border between the SAN and surrounding atrium [1]. The insulating border provides discrete SAN exit pathways that permit electrical coupling between the SAN and the atrial muscle. In addition, another anatomical feature that our laboratory have uncovered is a secondary pacemaker, named the paranodal area, in close proximity of the SAN [2]. The effects of these anatomical features on SAN electrical function have not been studied. Using our 3D anatomical model of the human SAN, we examined the significance of the anatomy in cardiac electrical wave dynamics. A functional 3D model of the human SAN was constructed using the detailed anatomy from our previous study [2]. The model consists of a SAN primary pacemaker surrounded by an insulating border. The insulating border provides discrete exit pathways to permit electrical coupling between the SAN and atrial muscle. The paranodal area is a column extending along the length of, but not in contact with, the SAN. Each tissue type was assigned excitation properties using validated variants of the Fenton-Karma cell model [3]. Parameter gradients within the SAN ensure that the leading pacemaker location is located at the centre of the SAN during physiological heart beats. Simulation experiments were performed using this dynamic model. In each experiment, the cases with and with the exit pathways as well as the cases with and without the paranodel area were simulated to permit comparision between various anatomical consutructions. Several simulation experiments involved re-entrant wave dynamics. To initiate the re-entry, the phase distribution method that we developed previously was exploited [4]. A spectrum of codes were implemented to permit data analysis. Our results show that cell-cell coupling gradients regulate the leading pacemaker location. The complex excitation initiation and propagation in the SAN region as observed experimentally [5] was possible by inclusion of the insulating border and could not be reproduced without the insulating border and exit pathways. When the SAN was made inactive, the paranodal area was capable of pacing the atrial muscle at a slower rate. The existence of an insulating border and discrete exit pathways promoted re-entrant wave breakup, and the arrythmic breakup was aggrevated by the presence of the spatially extended paranodal area. In addition, the paranodal and SAN oscillators may influence each others pacemaking propensity that may promote arrythmia. It was seen that the experimentally observed erratic activation of the SAN region could be reproduced in our simulations that had discrete exit pathways and paranodal area. Multiple re-entry based mechanisms could explain clinical atrial tachycardia.