Background. Monitoring of serum digoxin concentration (SDC) is one of the approaches for controlling the safety of the long-term digoxin treatment, however, the benefits from SDC monitoring in paroxysmal arrhythmias have not been studied. Aim. To evaluate the potential benefits of digoxin laboratory monitoring at the initiation of digoxin therapy in patients with paroxysmal atrial fibrillation (AF). Material and methods. A retrospective analysis of SDC was performed in patients with paroxysmal or persistent (n=142) and permanent AF (n=48) who received digoxin therapy. Anthropometric features, comorbidity, anamnesis of the underlying disease, concomitant therapy, severity of signs of myocardial remodeling and the presence of cardiac rhythm and/or conduction disorders before and after the treatment were considered to analyze the results of treatment. SDC was measured in 20 hours after digoxin treatment initiation and after restoration of sinus rhythm, or in 1 week after the start of therapy if arrhythmia persisted. Results. During the first week after digoxin therapy initiation the rate of non-specific ECG repolarization abnormalities was 54%, the incidence of clinically significant conductivity disturbances - 28%. We found a relationship between SDC in the first week of treatment and the risk of high grade AV- or SA-conduction abnormalities after the restoration of sinus rhythm (mean SDC 0.98±0.72 ng/dL in the group of patients with complications and 0.45±0.42 - in the group without them, p=0.015); these complications did not correlate with clinical signs of digitalization or intensity of digoxin therapy. The intake of digoxin was not an independent factor for the occurrence of cardiac rhythm disturbances. Comparison of the direct SDC measurements and the calculated values obtained with the valid calculators developed for the long-term digoxin use did not show reliable reproducibility of the calculated data. Conclusion. SDC monitoring may be used during the first week of therapy (before the establishment of a steady-state SD?) to minimize the risks of clinically significant adverse effects of digoxin on intracardiac conduction at the start of therapy. The existing empiric SDC calculators are not suitable as an alternative to the direct measurement. Further studies are needed to determine the clinically valid borderline values of the safe SDC in the condition of rapid digitalization.