The aim of the present study was to compare plasma levels of urokinase-type plasminogen activator (u-PA), before and after 20 min of venous stasis, with those of tissue-type plasminogen activator (t-PA), type 1 plasminogen activator inhibitor (PAI-1) and t-PA/PAI-1 complexes, to determine whether both plasminogen activators and their inhibitor respond similarly to the same stimulus. We studied 36, patients with recurrent venous thrombosis in whom no coagulation defects predisposing them to thrombosis had been detected (mean age 38.2 years, range 15–70 years). Twenty healthy individuals (mean age 34.3 years, range 20–60 years) served as a control group. t-PA, PAI-1 and u-PA activity and antigen, as well as the t-PA/PAI-1 complex antigen, were measured before and after venous stasis. Post-stasis fibrinolytic parameters were corrected for the haemoconcentration which occurred during the venous occlusion test. Pathologically high PAI-1 levels (antigen and activity) were found in four out of 36 patients who were excluded from study. Functional and antigenic u-PA increased significantly after venous stasis when analysed as the absolute differences between paired samples (P< 0.01). This increase in u-PA did not correlate with changes in t-PA or PAI-1 (r = 0.28 and r = 0.3fc respectively). This leads us to suggest that different mechanisms relating to clearance and/or release from diverse sources might be involved in elevations of u-PA in response to a local endothelial stimulus. We conclude that venous stasis might not be the elective choice when evaluating ‘bad responders' predisposed to thrombosis.