The sequential use of different types of antiarrhythmic drugs may improve arrhythmia prognosis in chronic atrial fibrillation or flutter after successful electrical cardioversion. The rationale for serial treatment is that the arrhythmogenic mechanism may vary between patients, leading to different responses to 1 specific drug. To investigate this issue prospectively, 127 patients having chronic fibrillation or flutter exclusively, underwent serial drug treatment with flecainide (stage I) followed by sotalol or, if contraindicated, quinidine (stage II) and eventually amiodarone (stage III). Stages II and III were entered after electrical recardioversion for a recurrence during stages I or II, respectively. Calculated on an actuarial basis, the 2-year cumulative percentage of patients free of the arrhythmia increased from 31% after stage I to 63% at the end of serial treatment. To reach this result, a mean of 1.8 +/- 0.8 cardioversions per patient were needed, with 53 patients progressing to stage II and 34 to stage III. Sixteen patients stopped serial treatment prematurely and 15 patients were considered to have intractable atrial fibrillation at the end of stage III. Incidence of proarrhythmia was low. Multivariate analysis disclosed that an older age, in combination with a large number of previous episodes of arrhythmia, a long previous duration of arrhythmia and presence of mitral valve disease, were predictive for medical refractoriness during serial treatment. It is concluded that serial treatment may improve arrhythmia prognosis in atrial fibrillation or flutter, with an acceptable incidence of proarrhythmic events.