A computer model of “modulated sinus parasystole” was devised in which two sinus pacemakers interacted electrotonically, entraining each other's periodicity according to their beat-tobeat phasic relationships. Depending on the preestablished rules, the model gave rise to various rhythm patterns that were similar to those recorded in patients with sinoatrial arrhythmias. The validity of the model in predicting clinically observed rhythm disturbances was tested in a case of sinoatrial extrasystolic activity. The sinoatrial origin of parasystolic discharges giving rise to various patterns of group beating in this case was diagnosed according to the following electrocardiographic criteria: (1) premature P waves having contour identical to P waves of basic beats, (2) variable coupling intervals, and (3) absence of compensatory pauses (i.e., returning cycles having duration similar to that of the basic P-P interval). For the analysis, it was assumed that two distinct but closely apposed sinoatrial pacemaker centers were competing for activation of the heart. The model accurately simulated the arrhythmias in the electrocardiographic trace. The best fit was found when the two pacemakers interacted on the basis of “resetting” in one direction and electrotonic modulation in the other. In fact, under appropriate conditions, the model matched precisely all frequency-dependent patterns of extrasystolic activity observed in the trace. We conclude that the modulated parasystole hypothesis can readily explain the mechanism of sinus extrasystolic discharges whose returning cycle equals the basic P-P interval. Moreover, the model predicts that, when the rules for mutual entrainment between "dominant" and parasystolic sinus pacemaker are appropriate, the retuming cycle can be shorter than the basic cycle.