This study examined whether an occlusive thrombus within a coronary artery can be lysed by streptokinase retroperfusion into the associated regional coronary vein. Experimental coronary artery thrombosis was induced in 15 closed chest dogs by placing a small copper coil at a proximal site of the left anterior descending coronary artery. Total thrombotic obstruction of this artery was verified within 10 to 60 minutes (38.0 +/- 15.8, mean +/- standard deviation) and streptokinase was administered within 94.0 +/- 17.4 minutes from coil insertion at an average rate of 42 IU/kg per minute by one of three modes: 1) intermittent 10 minute direct coronary venous retroinfusion (five dogs); 2) continuous infusion into the pumping circuit of synchronized phased retroperfusion of the great cardiac vein with arterial blood (five dogs); and 3) for comparison, streptokinase administered intravenously (five dogs). The intracoronary thrombus was fully lysed and anterograde reperfusion established within 51.0 +/- 18.7 minutes by intermittent streptokinase retroinfusion, and in 50.0 +/- 6.1 minutes by streptokinase supplemented synchronized retroperfusion (50.5 +/- 13.2 minutes for pooled retrograde coronary venous delivery). Lysis was also induced by systemic streptokinase, but the time to lysis was significantly longer and more variable (131.6 +/- 60.6 minutes) than with retrograde administration (p less than 0.01). The retroperfusion modality appears the preferable technique because it provides early thrombolysis and, at the same time, improves cardiac function and maintains myocardial viability of the jeopardized ischemic zone pending achievement of full reflow. Thus, streptokinase retroperfusion, if promptly instituted, may be a useful complemental nonsurgical treatment of evolving acute myocardial infarction after thrombotic coronary artery occlusion.