Possible correlation of the effects of pharmacotherapy on the inhibition of the in-vivo generation of thrombin and on the prevention of thrombus extension in patients with deep vein thrombosis (DVT) could help to define patients at higher risk. Patients with symptomatic deep vein thrombosis confirmed by phlebography were randomised to intravenous unfractionated heparin (UFH), or a subcutaneous low-molecular-weight heparin (reviparin) twice daily for one week, or a subcutaneous reviparin once daily for four weeks. The patients were treated with oral anticoagulants for at least 3 months. Main endpoints were regression of thrombus on phlebography on Day 21 and recurrent symptomatic venous thromboembolism up to 3 months. Coagulation parameters, markers of in-vivo thrombin generation, and TFPI-release were determined at randomisation, weeks 1 and 3. Four hundred sixty six responders (reduction of at least 30 per cent in Marder score) and 419 non-responders (Marder score unchanged or changed less than +/-30%) showed no significantly different baseline characteristics. The non-responder group had a higher median Marder score at baseline and after one and three weeks of treatment, and had significantly higher fibrinogen levels, TAT complexes and F1+2 values than responders. There were no significant differences in coagulation parameters between non-responders and patients with asymptomatic + symptomatic VTE with the exception of higher TAT complexes at baseline. Significant differences in Marder score and coagulation parameters at baseline were found between responders and nonresponders. Non-responders have a higher risk tosuffer recurrent VTE and may need intensified treatment.