Background and Aim: The sinoatrial node (SAN) is electrically and structurally heterogeneous [1, 2]. SAN cell apoptosis may accentuate the heterogeneity causing SAN exit block. Methods: Our previous model [3] was enhanced by including a histology based 2D mouse SAN geometry [2]. A semi-insulating SAN-atrial boundary was randomly assigned a number of SAN-atrial exit pathways. A simple electrophysiological cell model simulated a range of SAN pacemaking firing rates, as well as atrial excitable action potentials (APs) (see Figure). The spatial distribution of SAN cells was assigned randomly. Gap junction coupling (GJC) in the atrium was adjusted to produce a 0.6 m/s conduction velocity. SAN GJC distribution was made heterogeneous. Maximum SAN GJC was 10 times smaller than in the atrium. Inactive apoptotic cells and the insulating tissue were assigned a small electrical conductivity. Each simulation produced 5 s of electrical activity data. To detect the SAN exit block, the number of oscillations at each location were recorded. Bradycardia was defined as when number of SAN oscillations exceeded those in the atrium. Standard deviation of each SAN cell’s oscillations identified the leading pacemaker locations [4, 5]. The estimated leading pacemaker locations was verified by studying time animations of the 2D voltage distribution. Three simulation experiments were performed. In the first experiment, the number and size of exit pathways was gradually increased till a robust pacemaking-propagation pattern was acheived. This estimate of the exit pathway density along the insulating border was used in the next two experiments. In the second experiment, the occurance of SAN exit block due to uniformly ditributed apoptotic cells (0% and 10%) was tested. The third experiment was similar to the second experiment, except that the apoptotic cells were confined to be within 0.2 mm of the insulating border. Results: The outcome of the first experiment indicates that the amount of exit pathways should be 17.5% to acheive a robust SAN-atrial pacing. The leading pacemaker location was found to coincide with the location of minimum GJC in the SAN. The results of simulations 2 and 3 show that with progressive apoptosis, SAN-atrial pacing was increasingly intermittent. Multiple distinct leading pacemakers separated by 0.1 to 0.3 mm were observed. Whereas a total arrest of atrial pacing was observed at 8% in simulation 2 (uniformly distributed apoptosis) [3], total arrest was observed at a lowever 4% apoptosis in simulation 3 (apoptosis close to SAN-atrial boundary). Conclusions: The model behaviour indicates that only a small part of the SAN has to be in contact with the atrium to acheive physioloical pacing. In contrast to our previous study, the new model is capable of reproducing the experimental data of missed beats and sinus block at modest levels of apoptosis.