Introduction: It has been shown that repolarization alternans (Re-ALT), a beat-to-beat alternation in action potential duration, enhances dispersion of repolarization DOR above a critical heart rate and promotes susceptibility to ventricular arrhythmias. It is unknown whether Re-ALT is measurable in the atria using standard pacemakers and whether Re-ALT plays a role in promoting AF. Methods: two DDD pacemakers (PM) with two right atrial (RA) leads were implanted in 3 male sheep (>60kg) after ablation of the AV junction. The 1st PM was used to deliver rapid pacing (400 beats) for measurements of RA Re-ALT, the 2nd one to record a broadband unipolar RA electrogram (EGM). Re-ALT was considered present when the 2 following conditions were fulfilled: 1) beat-to-beat alternation in T-wave apex amplitude (Ta) for >4 consecutive beats, and 2) statistically different distributions of odd and even Ta (p<0.01). Results: Re-ALT amplitude progressively increased as a function of pacing CL from a mean value of 52±5 μV at 400ms to a maximal value of 184±36 μV at a mean pacing CL of 210±36 ms. Importantly, Re-ALT appeared intermittent but no periodicity was noticed, and its onset was independent of the pacing CL. In one sheep, rapid pacing at a CL of 230 ms induced an episode of non sustained AF that was preceded by overt Re-ALT. Conclusions: in-vivo measurement of atrial Re-ALT shows that Re-ALT amplitude increased as a function of pacing rate. Re-ALT appeared intermittent but no periodicity was observed. The observation of overt Re-ALT preceding an episode of non sustained AF suggests that Re-ALT is a promising parameter for in vivo assessment of AF susceptibility using standard pacemaker technology.