Summary. Background: The assessment of the risk of recurrent venous thromboembolism (VTE) is important to determine the optimal duration of secondary prophylaxis. The risk can be estimated by measuring individual parameters reflecting hypercoagulability. Because of the large numbers of such putative parameters, the assessment in individual patients is complex. Application of global assays reflecting the pro-/anti-coagulant balance in vivo would be desirable. Objectives: To investigate the relationship between recurrent VTE and thrombin generation (TG). Patients-methods: Two hundred and fifty-four patients were followed-up after a first episode of unprovoked, objectively documented VTE for a period of 2.7 years after discontinuation of treatment with vitamin K antagonists. TG was measured 1 month after discontinuation of treatment as endogenous thrombin potential (ETP), peak thrombin and lag-time in the presence or absence of thrombomodulin. The study outcome was objectively documented symptomatic recurrent VTE. Results: Patients with ETP or peak (measured in the presence of thrombomodulin) of >960 nm*min or >193 nm had hazard ratios (HR) (95% CI) for recurrent VTE of 3.41 (1.34–8.68) or 4.57 (1.70–12.2) as compared with those with an ETP <563 nm*min or peak <115 nm. Patients with lag-time <14.5 min had HR of 3.19 (1.29–7.89) as compared with those with lag-time >20.8 min. HR for ETP, peak or lag-time measured in the absence of thrombomodulin were smaller than those measured in the presence of thrombomodulin. Conclusions: The measurement of TG helps to identify patients at higher risk of VTE recurrence. The increased risk may be better appreciated if the test is performed in the presence of thrombomodulin.