AIMS: To assess the feasibility of placing permanent atrial pacing leads during atrial fibrillation (AF) and whether such leads function satisfactorily. METHODS AND RESULTS: Prospective study of 17 consecutive patients in whom permanent atrial leads were positioned during an episode of paroxysmal AF. Fluoroscopic position ('figure of 8' or side-to-side movement and anterior position in RAO projection), lead impedance (> 300 but <1000 ohms) and intracardiac electrogram (average peak to peak amplitude > 1 mV) were used to define an acceptable lead position. At 8 weeks post implant we measured: pulse duration pacing threshold at 5 V; lead impedance at 5 V and 0.5 ms; intracardiac electrogram (EGM) signal amplitude. At the end of the study we reviewed patients to establish whether AF had become permanent. In all patients, follow-up demonstrated satisfactory lead function. All leads had impedances between 300 and 1000 ohms. Pacing thresholds were all <0.1 ms at 5 V. Mean atrial EGM amplitude seen in sinus rhythm was 3.3 mV (range 1.2-8.4); in patients where all follow-up was in AF in was 2.1 mV (range 1.5-2.5). Nine patients (53%) developed permanent AF. CONCLUSION: Placing atrial leads during AF is feasible using the technique described. However, some patients progress to chronic AF, eliminating the benefits of atrial pacing.