Emphasis on the role of patent foramen ovale as a potential risk factor for ischemic paradoxical stroke has recently increased. Current therapeutic options for secondary stroke prevention include long-term antithrombotic therapies and invasive closure of the defect, but selective indications have not been evaluated. Therefore we developed a Markov-based decision analysis model for a hypothetical cohort of patients 55 years of age with presumed paradoxical embolism, measuring for each therapy the risks of stroke recurrence, treatment-related complications, and death after 5 years and the quality-adjusted life-years. Over a wide range of stroke risk recurrence (0.8% per year to 7% per year), the gain provided by closure of the defect exceeded the one obtained by other therapeutic options. When the risk exceeded 0.8% per year and 1.4% per year, respectively, this was also verified for anticoagulation and antiplatelet therapies compared with therapeutic abstention. Therapeutic abstention was the preferred strategy under 0.8% per year. Sensitivity analyses identified key parameters influencing the choice of therapy. These included estimates of stroke recurrence, bleeding rates, surgery-related case fatality rates, and age. Considering the risks of treatment and the devastating consequences of a recurrent stroke, our model suggests that if the estimated risk of paradoxical stroke recurrence is >0.8% per year, therapeutic abstention becomes the worst option. Above this threshold secondary stroke prevention with anticoagulation therapy or surgical closure of the defect is the preferred strategy, and assessment of both the risk of stroke recurrence and the risk related to therapeutic options should guide individual therapeutic decision making. (Am Heart J 1998;135:532-41.)