Myocardial distribution of cardioplegic solution infused by combined antegrade/retrograde routes was assessed with myocardial contrast echocardiography in 18 patients with chronic stable angina and three-vessel disease undergoing elective coronary artery bypass grafting. Overall myocardial opacification was significantly greater in retrograde than in antegrade cardioplegia (77.7% +/- 13.4% versus 59.1% +/- 15.7%; p = 0.0009). The difference was affected by collateral circulation, as pointed out by the significant interaction between coronary collateral circulation and percent of myocardial opacification after antegrade and retrograde cardioplegia (p = 0.002). When we performed multiple comparisons, in patients with good collaterals the opacification difference between antegrade and retrograde cardioplegia was not statistically significant (66.4% +/- 10.2% versus 76.0% +/- 15.2%; p = not significant), whereas in patients with poor collaterals myocardial opacification during retrograde cardioplegia was significantly greater (44.3% +/- 15.0% versus 81.2% +/- 9.0%; p < 0.02). During antegrade cardioplegia, patients with poor collaterals showed a lower degree of myocardial opacification than patients with good collaterals (44.3% +/- 15.0% versus 66.4% +/- 10.2%; p < 0.01). Our results show that retrograde cardioplegia in patients undergoing elective coronary artery bypass grafting offers no advantage over antegrade cardioplegia when collateral circulation is well developed. On the other hand, conventional aortic root infusion may not provide adequate myocardial protection in the subset of patients with significantly narrowed or occluded coronary arteries and poor collaterals.