In subjects without coronary disease, coronary perfusion pressure generated with closed-chest cardiopulmonary resuscitation (CPR) bears a direct relationship to myocardial blood flow. The effect of coronary lesions on this relationship was studied in an experimental porcine model not requiring thoracotomy. Coronary stenoses (a 50% reduction in coronary cross-sectional area) or total coronary occlusions were created by percutaneous, transarterial catheter placement of a Teflon cylinder in the left anterior descending artery of 21 swine (30 to 60 kg). Coronary perfusion pressure, defined as the aortic diastolic pressure minus right atrial diastolic pressure, was correlated with myocardial blood flow measured with nonradioactive, colored microspheres during external chest compression CPR. Complete occlusion of the left anterior coronary artery resulted in essentially no CPR-generated blood flow to the anterior myocardium distal to the site of occlusion. Coronary perfusion pressure showed a positive correlation with myocardial blood flow above the area of occlusion (r = 0.783; p less than 0.01) but did not correlate with myocardial blood flow below the occlusion site (r = 0.239). In the presence of a patent coronary artery stenosis, coronary perfusion pressure correlated with myocardial blood flow both above (r = 0.841; p less than 0.001) and below (r = 0.508; p less than 0.05) the stenosis. During closed-chest CPR producing coronary perfusion pressures between 30 and 60 mm Hg, anterior myocardial blood flow was 109 +/- 16 ml/min/100 gm above a patent stenosis and 66 +/- 13 ml/min/100 gm below the stenosis (p less than 0.005). Over a wide range of coronary perfusion pressures, myocardial blood flow below a coronary lesion was significantly less than that above the lesion. Coronary occlusions and stenoses can substantially affect the amount of CPR-generated coronary perfusion pressure needed to produce distal myocardial blood flow.