Whether to continue oral anticoagulant therapy indefinitely after completing 3 to 6 months of oral anticoagulant therapy for "unprovoked" venous thromboembolism (VTE), is one of the most important unanswered questions in VTE management. This long-term decision should be based on balancing the long-term mortality risk from recurrent VTE, largely preventable with oral anticoagulant therapy, against the long-term mortality risk of major bleeding, the principle complication of oral anticoagulant therapy. There exist important knowledge gaps in estimating the long-term mortality risk of recurrent VTE in patients with unprovoked VTE who discontinue therapy and the long-term mortality risk from major bleeding in those who continue oral anticoagulant therapy. These knowledge gaps, reviewed herein, are the source of uncertainty for patients and health care providers wrestling with this important question. One promising solution is recurrent VTE risk stratification where unprovoked VTE patients are categorised as low or high risk for recurrent VTE and clinical decision making is less ambiguous and ultimately will likely lead to better outcomes.