Diagnostic tests are a cornerstone in modern medicine. They are used not only to confirm the presence of a disease but also to rule out the disease in healthy subjects. Tests with two outcome categories (i.e. presence/absence) are known as dichotomous tests. Their inherent validity is determined by sensitivity and specificity and the receiver operating characteristic (ROC) curve is known to be a simple, yet complete plot that displays the full picture of trade-off between the sensitivity (true positive rate) and (1- specificity) (false positive rate) across a series of cut-off points. Our study found that, even in the early hours of paroxysmal atrial fibrillation, there were significant changes in major indicators of fibrinolysis, namely plasminogen level, t-PA level, PAI-1 activity, α2-antiplasmin activity, vitronectin and D-dimer plasma levels. We believe that they are closely related and stem from the disease itself. This gave us reason, using these indicators as predictors, to search for a diagnostic option to rule out PAF. We used statistical models of logistic regression analysis and ROC to achieve this. Values of p<0.05 were considered statistically significant. Plasma levels of vitronectin have been found to be the most reliable predictor for ruling out PAF (specificity 88%, sensitivity 83%, AUC 0.96), while D-dimer levels had the lowest diagnostic values (37% specificity, 81% sensitivity, AUC 0.56). The obtained results are not only of pure scientific but also of applied nature. They could be used to improve identification of patients at risk for PAF embolism, and assist in the choice of thromboprophylaxis.