The long-term aims of treatment of venous thromboembolism (VTE) are the prevention of late recurrences and the post-thrombotic syndrome. The optimal duration of oral anticoagulant therapy (OAT) after VTE is still controversial. Patients can be stratified into risk categories for recurrence on the basis of clinical characteristics of the index VTE event. Patients with VTE due to a transient risk factor are at a low risk of recurrence and short term anticoagulation is indicated (3 months). Patients with an idiopathic event or with known thrombophilic defects are candidates to a longer OAT course. Patients with cancer, antiphospholipid antibodies syndrome, recurrent idiopathic event, antithrombin deficiency, homozygosity or double heterozygosity for factor V Leiden or prothrombin mutation are candidates for extended long-term anticoagulation. More recently, studies have indicated that other factors such as D-Dimer levels after the discontinuation of OAT or the residual vein thrombosis could be additional predictive factors for recurrences. The recent Italian collaborative, prospective, randomized PROLONG study showed that patients with a previous idiopathic VTE, treated for at least 3 months, who have abnormal D-dimer assay tested one month after OAT withdrawal have a significant incidence of recurrent VTE which is reduced by resumption of anticoagulation. On the other hand, the optimal course of anticoagulation in subjects with normal D-dimer, whose risk of VTE recurrence is markedly lower, is not clearly established yet.