Background. Digoxin takes important place in the management of patients with paroxysmal or persistent atrial fibrillation (AF). Monitoring of serum digoxin concentration (SDC) seems to be perspective way to improve the safety and efficacy of treatment. At the same time, there are no generally accepted reference ranges for SDC, especially in terms of the onset of therapy Aim. To evaluate the potential contribution of SDC monitoring in the context of efficacy of AF treatment with the use of digoxin in elderly patients. Material and methods. A retrospective analysis of treatment of patients with recent-onset (n=91) or permanent (n=58) AF was performed. In all cases, the strategy of heart rate (HR) control was realized, including treatment with digoxin in 104 cases. SDC was measured twice during the digitalization: 20 hours after the first digoxin dose, and at the endpoint (time of sinus rhythm spontaneous restoration or, if the rhythm not restored, on day 7 of AF persistence). Results. The influence of digoxin on HR was dose-dependent since the 1st week of therapy, and the SDC dynamics was strictly associated with change in HR (r=-0.66, p<0.001). There was a negative correlation between the probability of the sinus rhythm restoration due to digoxin therapy and the SDC: its probability in high SDC was significantly lower compared to other approaches to the HR control (0% vs 76% in beta-blocker monotherapy; and vs 100% in therapy with beta-blocker + lower-SDC, p=0.036). In patients with persisted AF and low SDC, the digoxin was withdrawn more often due to lack of clinical efficiency Conclusions. The SDC monitoring at the 1st week of digitalization could be used to improve the efficacy of therapy and to minimize the risks of spontaneous rhythm restoration, if not desired, and to reveal the disproportion between its high dose and low efficacy. The role of this diagnostic tool seems to be limited to safety control and negative predictive value for efficacy (the lower concentration, the higher risk of inefficiency), whereas its positive predictive value in terms of efficacy seems contradictory. The obtained data could be used for decision-making for recommendation of longterm digoxin usage if its contribution into HR control is doubtful.