This study tests the hypothesis that retrograde/antegrade cardioplegic delivery can overcome the limitations of poor cardioplegic distribution resulting from either technique alone and, potentially, may expand the safety of using internal mammary artery grafts in cardiac muscle in jeopardy of inadequate cardioplegic protection. Jeopardized myocardium was produced in 20 dogs by ligating the left anterior descending coronary artery for 15 minutes before starting cardiopulmonary bypass and by 1 hour of aortic clamping with multidose 6 degrees C cold blood cardioplegia. Five dogs received antegrade cardioplegia via the aortic root. Ten dogs received retrograde cardioplegia via the coronary sinus. Five additional dogs received retrograde/antegrade cardioplegia via both routes. The ligature on the left anterior descending coronary artery was removed after aortic unclamping, and regional myocardial temperature (thermistor probe), segmental shortening (ultrasonic crystals), and global left ventricular and right ventricular myocardial function were evaluated. Antegrade cardioplegia produced excellent right ventricular cooling (14 degrees C) and allowed complete right ventricular functional recovery. However, it failed to cool muscle supplied by the left anterior descending coronary artery (only 31 degrees versus 12 degrees C, p less than 0.05), postischemic global left ventricular function recovered only 38% (p less than 0.05), and segmental shortening in the region supplied by the left anterior descending coronary artery recovered only 22% (p less than 0.05). Retrograde cardioplegia produced homogeneous cooling (17 degrees C) and allowed near normal recovery of global and regional left ventricular function (99% and 86%), but right ventricular cooling was variable (19 degrees to 30 degrees C) and right ventricular function recovered inconstantly (range 64% to 100%, average 82%). The best myocardial protection occurred after retrograde/antegrade cardioplegia; myocardial cooling was homogeneous, left ventricular and right ventricular global function recovered completely (95% and 90%), and regional contractility in muscle supplied by the left anterior descending coronary artery returned to 84% of control. We conclude that retrograde/antegrade cardioplegia provides better myocardial protection than either technique alone, ensures good cardioplegic distribution to the left and right ventricles, and allows regional delivery of cardioplegic flow to segments supplied by occluded arteries.(ABSTRACT TRUNCATED AT 400 WORDS)