Atrial septal (Se-P) and atrial appendage pacing (Ap-P) were compared in a randomized, controlled study to assess the feasibility, the reliability, and the effects of Se-P on atrial conduction, interatrial synchronization, and the AV sequence. The main baseline characteristics of the patients were comparable in both groups. There was no difference in feasibility or reliability between the two techniques. Compared to Ap-P (n = 28), Se-P (n = 28) decreased the P wave duration, left atrial electromechanical delay (LAEMD), and interatrial interval (-1.6% vs +28%, P < 0.001; -3% vs +30%, P < 0.001; -130% vs +78%, P < 0.001); it induced a smaller increase of the right AEMD, a slight reversal of the timing of the atrial systoles and a shortening of the PR interval (-13% vs +25%, P < 0.001) and of the interval separating atrial systoles from ventricular activation. Finally, the shortening of the PR interval was smaller during high Se-P versus low Se-P. Se-P avoids the undesirable prolongation of the atrial, interatrial, and AV conductions observed during Ap-P. In addition, Se-P creates a slight reversal of the timing of the atrial systoles and induces a shortening of PR interval, the extent of which could depend on the height of the pacing site on the septum.