Abstract Background QT interval varies with the heart rate (HR), so a correction in QT calculation is needed (QTc). Atrial fibrillation (AF) is associated with elevated heart rate (HR) and beat-to-beat variation in the length of the QT and RR interval making correction of the QT interval challenging. Aim to evaluate which correction formula better correlates with QTc in sinus rhythm after electrical cardioversion (ECV). Methods During a 3-month period, we considered patients who underwent 12-lead ECG recording for standard clinical indications and received an AF diagnosis. Exclusion criteria were the following: a QRS duration >120 ms, chronical therapy with QT prolonging drugs, a rate control strategy, a non-electrical cardioversion. The final population was composed by 60 patients. The QT interval was measured in V2 lead as the 10-beat average and corrected using Bazzett's, Framingham, Fridericia and Hodges formulas during the last ECG taken during AF and the first one immediately after ECV. Moreover, QTc mean was calculated as mQTc (average of 10 QTc calculated beat per beat) and as QTcM (QTc calculated from the average of 10 raw QT and RR for each beat). Results Population was composed by 26 women (43%) with a mean age was 79 ± 11 years. Good correlations between mQTc and QTcM are present for each formula, even in atrial fibrillation or sinus rhythm. Comparing atrial fibrillation and sinus rhythm a significant change in HR was seen (63,9 ± 14,9 vs 99,5 ± 27,9, p<0,001). Bazett's formula shown a significant change in mean QTc value between the two rhythms (421,5 ± 33,9 vs 446,1 ± 31,9; p< 0,001 for mQTc and 420,9 ± 34,1 vs 441,8 ± 30,9; p = 0.003 for QTcM). On the contrary, QTc assessed by the Framingham, Fridericia, and Hodges formulas was similar to that in AF (Figure). Conclusion In ECG during atrial fibrillation, due its high heart rate, Bazzett's formula, although the most used, seems to be the most imprecise in QTc estimation.