Atrial mechanical dysfunction after cardioversion for atrial fibrillation has been widely evaluated in recent years. Nevertheless, the influence of many clinical and echocardiographic parameters is not yet understood. The aim of the present study was to evaluate the influence of clinical and echocardiographic parameters on the return of effective atrial contraction. A total of 109 patients were evaluated: 41 patients had spontaneous recovery of sinus rhythm and 68 patients were randomly treated using either direct-current (DC) shock or intravenous procainamide. Elective cardioversion was accomplished pharmacologically in 23 patients (67%) and with DC shock in 29 patients (85%). Patients underwent a complete echocardiographic examination 1 hour after the restoration of sinus rhythm and after 1 and 7 days and 1 month. The following parameters were evaluated: patient age, cardiac disease, duration and etiology of atrial fibrillation, mode of cardioversion, left ventricular diameters and function, and left atrial diameter and function assessed as atrial ejection force. The relation between these variables and atrial ejection force was tested. Atrial ejection force was greater immediately and 24 hours after cardioversion in patients who had spontaneous recovery of sinus rhythm and in patients treated with drugs than in patients treated with DC shock. The mode of cardioversion was significantly associated with the recovery of atrial mechanical function by day 1 in univariate and multivariate analyses (odds ratio 0.14; 95% confidence interval 0.02 to 1.2). The other variable associated with the recovery of function was normal left atrial size (odds ratio 0.16; 95% confidence interval 0.12 to 1.6). In conclusion, atrial ejection force is a noninvasive parameter that can be easily measured and can provide accurate information about the recovery of left atrial mechanical function. The recovery of atrial function was influenced by the mode of cardioversion and the size of the left atrium.