Vascular plasminogen activator was measured by means of a new chromogenic assay in 24 women who had suffered from oral contraceptive-associated thrombotic disease and was compared to that in a control group of 78 premenopausal women. Vascular plasminogen activator levels were significantly reduced in the subjects who had venous thrombosis but not in the five women who had arterial thrombosis (0.04 +/- 0.03 versus 0.38 +/- 0.31, respectively) when compared to the levels in the control group (0.19 +/- 0.20). Since vascular activator levels distribute in a non-Gaussian manner, cases and controls were also stratified into deciles. Seventeen subjects who had suffered from venous thrombosis were stratified in the lowest three deciles, and two subjects, in the fourth and fifth deciles. Subjects who had suffered from arterial thrombosis were in the fourth or higher deciles. The conclusion is that, although there is a correlation between low levels of vascular plasminogen activator and venous thrombosis, no such correlation exists for arterial thrombosis.