We utilized low-dose intraembolic urokinase (UK) infusions in a canine model of experimental pulmonary embolism (PE) and compared the arteriographic extent of thrombolysis with three other treatment regimens. Group 1 animals (n = 16) received the intraembolic UK infused directly into the PE offering the mechanical effect of the infusion combined with pharmacologic thrombolysis. In the group 2 animals (n = 5), UK was delivered via a guide catheter placed proximal to the clot. Group 3 animals (n = 6) were treated with a direct intraembolic saline solution infusion. Group 4 (n = 7) received only intravenous heparin. The arteriographic extent of thrombolysis was graded 1+ to 3+. The extent of thrombolysis was 2.88+ in the group 1 animals and was significantly greater than in groups 2, 3, or 4 (p = 0.003, 0.0005, and 0.0001, respectively). Fibrinogen levels did not significantly decrease with intraembolic treatment (p = 0.07). Delivery of UK directly into emboli in an experimental canine PE model appears to elicit a combined mechanical and pharmacologic effect resulting in extensive thrombolysis.