Making decisions about any modality of secondary prophylaxis in patients with venous thromobembolism (VTE) has to balance the risk of bleeding induced by anticoagulants against the benefit of reducing the risk of recurrent disease. It has to be kept in mind that the magnitude of risk is not only defined by the number of events per time period but also by the impact of the event on the fate of the patient. With standard intensity vitamin K antagonists (VKA), the risk of bleeding is more closely related to comorbidities than to other factors, eg age. The risk of VTE recurrence differs largely between patient groups. The criterion of presence or absence of a permanent or transient clinical trigger factor for the actual VTE episode has a greater impact than an abnormal result in thrombophilia testing. The standard period of secondary prophylaxis for proximal DVT and for PE is three to six months. The concept of prolonging this period for several months according to the risk of recurrence is seriously challanged by the observation that the prolongation period seems to delay recurrencies rather than truly avoiding them. For this reason, patients who clearly are threatened by recurrent episodes should receive indefinitive secondary prophylaxis. This is the case for cancer patients, patients with the antiphospholipid syndrome, and those who belong to families with severe and symptomatic protein C, protein S, or antithrombin deficiencies. Patients with recurrent VTE, with idiopathic VTE, or with combined thrombophilic conditions may only benefit from indefinitive secondary prophylaxis if the bleeding risk of the anticoagulant regimen under consideration is very low.