During coronary angioplasty, inflation of the balloon within the coronary artery produces transient arterial occlusion and frequently results in myocardial ischemia. Delivery of oxygenated autologous blood to the myocardium at risk during inflation may help mitigate this ischemia. Accordingly, we investigated the feasibility and safety of infusing blood through the central lumen of a dilatation catheter around the guidewire using both a model in vitro and clinical trials. In the tests in vitro, fresh blood was infused at flow rates up to 120 ml/min. Hemolysis was minimal at flow rates of 60 ml/min or less (≤ 0.92 ± 0.18%), but increased exponentially at higher rates (13.64 ± 2.37% at 120 ml/min, p < .002). A similar pattern was observed for potassium release. Platelet and leukocyte counts did not vary significantly, and β-thromboglobulin and muramidase remained at control levels. Although mean erythrocyte volume did not change, erythrocyte histograms and light microscopy demonstrated a subpopulation of red cell fragments averaging 25 to 40 fl in size at higher rates. A randomized, crossover clinical trial was next performed by delivery of blood perfusion at 60 ml/min to 15 patients undergoing coronary angioplasty. Levels of plasma hemoglobin, β-thromboglobulin, lactate dehydrogenase, and potassium remained constant before and after the perfusion and the control inflations. The maximum pain score was significantly lower with the perfusion inflation (4.1 ± 0.8 vs 6.0 ± 0.9, p < .003). Relative to baseline, the maximum ST segment elevation during the perfusion inflation (0.5 ± 0.3 mm) was nearly one-fourth that during the control inflation (1.9 ± 0.6 mm, p < .02). Thus, myocardial protection with oxygenated autologous blood perfusion at rates of 60 ml/min appears to be a safe and effective technique that may permit increased inflation time and extend the range of coronary angioplasty to include individuals at high risk for the procedure.