Normal atrial mechanic function may not return immediately after the successful cardioversion of atrial fibrillation. It has been suggested that the delayed recovery of atrial contraction (atrial stunning) might be due to: 1. the energy delivered during direct current cardioversion 2. the time from the onset of atrial fibrillation 3. the left atrial size 4. the associated cardiac disease. This study evaluates "atrial stunning" in patients pharmacologically treated, with atrial fibrillation of recent onset, normal atrial size and without heart disease. Doppler echocardiography is well suited for assessment of atrial function due to the ability of recording the peak velocity of atrial contraction (A wave).Twenty-five patients with no evidence of heart disease and M-mode left atrial dimension less than 40 mm underwent successful pharmacologic cardioversion (pro-paphenon or flecainide 2 mg/kg/10 min) of atrial fibrillation of recent onset (less than 48 hours). After cardioversion an echocardiographic study was performed within 12 hours (ECO 1), on day 3 (ECO 2), on day 12 (ECO 3), and on day 30 (ECO 4).No significant difference of both left atrial size (37 +/- 3.9 mm; 38.22 +/- 3.8 mm; 38.02 +/- 4.7 mm; 38.2 +/- 4.14 mm) and peak E velocity (57.97 +/- 18.3 mm/sec; 59.4 +/- 18.3 mm/sec; 59.0 +/- 16 mm/sec; 59.07 +/- 16.7 mm/sec) was demonstrated among serial echocardiographic evaluations. Both peak A velocity (mm/sec) and E/A ratio were significantly different in ECO 1 (60.29 +/- 12.3-1.0 +/- 0.37) than in ECO 2 (73.1 +/- 10.7, p < 0.005-0.82 +/- 0.27, p < 0.05); no statistical difference was found between ECO 2 and ECO 3 (76.31 +/- 12-0.78 +/- 0.24 mm/sec)--ECO 4 (76.91 +/- 14.8-0.78 +/- 0.21 mm/sec).This study suggests that patients with atrial fibrillation of recent onset have a delayed recovery of normal atrial systolic function after pharmacologic cardioversion.