The aim of the study was the development and clinical application of patient selection algorithm for surgical treatment of longlasting persistent atrial fibrillation. The study included 235 patients with acquired heart disease and coronary artery disease, which in the period from 1999 to 2015 performed surgical treatment of long-term persistent atrial fibrillation (RF “MAZE III” procedure) in conjunction with the correction of the underlying heart disease. The patients were divided into 2 groups according to the method of operation: the group 1 – 135 patients (76 women and 59 men) who have applied an integrated approach to surgery for atrial fibrillation, including penetrating method of RF effects on atrial myocardium and the study of the function of the sinus node before and after the operation (these patients were operated on from 2008 to 2015). The group 2 – 100 patients (62 women and 38 men) with a “classical” method of monopolar RF “MAZE III”, which the sinus node function was not studied. We used the combined (epi- and endocardial) method of RF "MAZE". This algorithm is decreasing of possible permanent pacemaker postoperatively. The initial sinus node function in these patients, measured using the original method, the basic line of this algorithm was taken. The results showed that use this algorithm for selection of patients allows significantly reduce the possibility of pacemaker implantation in the postoperative period