Introduction: Sophisticated monitoring of atrial activity is a prerequisite for modern pacemaker therapy. Ideally, near-fields and ventricular far-fields ought to be distinguished by beat-to-beat template analysis of the atrial signal. A prerequisite is that atrial signals are stable under different conditions.Methods and Results: A Matlab routine was developed to analyze atrial electrograms of 23 patients at least 3 months after implantation of a dual chamber pacemaker under several conditions including at rest, bipolar at rest, in an upright position, during treadmill exercise, and postexercise. A near-field and far-field template was created and amplitudes, widths, and slew rates were measured. In bipolar configuration, near-field amplitude at rest was 3.04 ± 0.94 mV (unipolar)/3.36 ± 1.0 mV (bipolar) versus 3.18 ± 1.0 mV (bipolar) at peak exercise. Far-field amplitude at rest was 1.66 ± 1.18 (unipolar)/0.47 ± 0.27 mV (bipolar) and 0.41 ± 0.21 mV (bipolar) at peak exercise (n.s. for bipolar measurements). No overall significant changes were observed for near-and far-field widths and slew rates during exercise. Shorter tip-ring distances of the atrial bipole, lead position, and the presence of sinus node disease did not have any impact on overall near-and far-field signal characteristics. Intraindividual differences between rest and peak exercise were moderate (range: near-field +0.15 to −0.54 mV; range: far-field +0.05 to −0.18 mV).Conclusions: Atrial near and far fields can be automatically classified and quantified by automated signal processing. Signals did not change during exercise or change of posture. This is a prerequisite for the implementation of beat-to-beat template analysis into pacemakers.