A technique of quantitative venography has been developed in which values are assigned to the deep veins of the calf, knee, thigh, and pelvis, based upon the calculated volume and degree of occlusion of these venous segments. A maximum score of 40 units reflects complete thrombosis of all segments. This technique has been applied to a randomized, single-blind study of streptokinase versus heparin treatment. Each group of 12 patients had similar mean inital venographic scores; follow-up venograms were performed 5 days after the start of therapy. Streptokinase patients with high initial scores (larger than 20) showed a mean improvement of 12.1 units, while those with low initial scores(less than 20) were essentially unchanged. Heparin patients with high scores had a minimal mean improvement of 1.1 units, but those with low scores had a significant mean extension of thrombosis of 8.6 units. Patients with symptoms for less than 7 days showed greated mean improvement (12.7 units) with streptokinase that those with a longer duration of symptoms (2.0 units); heparin patients in these subgroups showed a mean worsening of 7.5 units and no change, respectively. Extrinsic venous obstruction by tumor did not prevent an excellent response to streptokinase. No single test of coagulation of fibriolysis was a reliable indicator of the degree of venographic response to lytic therapy. Pyrexia and hemorrhagic complications occurred in over one-half of the streptokinase patients; one had an anaphylactic reaction, and one died of intracerebral hemorrhage during therapy. The data suggest that lytic therapy is best restricted to the patient with acute extensive thrombosis. Also, continuous infusions of heparin according to current guidelines may be inadequate to prevent thrombus growth in some patients.