To assess the likelihood of procedural success in patients with multivessel coronary disease undergoing percutaneous coronary angioplasty, 350 consecutive patients (1,100 stenoses) from four clinical sites were evaluated. Eighteen variables characterizing the severity and morphology of each stenosis and 18 patient-related variables were assessed at a core angiographic laboratory and at the clinical sites. Most patients had Canadian Cardiovascular Society class III or IV angina (72%) and two-vessel coronary disease (78%). Left ventricular function was generally well preserved (mean ejection fraction, 58 +/- 12%; range, 18-85%) and 1.9 +/- 1.0 stenoses per patient had attempted percutaneous coronary angioplasty. Procedural success (less than or equal to 50% final diameter stenosis in one or more stenoses and no major ischemic complications) was achieved in 290 patients (82.8%), and an additional nine patients (2.6%) had a reduction in diameter stenosis by 20% or more with a final diameter stenosis 51-60% and were without major complications. Major ischemic complications (death, myocardial infarction, or emergency bypass surgery) occurred in 30 patients (8.6%). In-hospital mortality was 1.1%. Stepwise regression analysis determined that a modified American College of Cardiology/American Heart Association Task Force (ACC/AHA) classification of the primary target stenosis (with type B prospectively divided into type B1 [one type B characteristic] and type B2 [greater than or equal to two type B characteristics]) and the presence of diabetes mellitus were the only variables independently predictive of procedural outcome (target stenosis modified ACC/AHA score; p less than 0.001 for both success and complications; diabetes mellitus: p = 0.003 for success and p = 0.016 for complications). Analysis of success and complications on a per stenosis dilated basis showed, for type A stenoses, a 92% success and a 2% complication rate; for type B1 stenoses, an 84% success and a 4% complication rate; for type B2 stenoses, a 76% success and a 10% complication rate; and for type C stenoses, a 61% success and a 21% complication rate. The subdivision into types B1 and B2 provided significantly more information in this clinically important intermediate risk group than did the standard ACC/AHA scheme. The stenosis characteristics of chronic total occlusion, high grade (80-99% diameter) stenosis, stenosis bend of more than 60 degrees, and excessive tortuosity were particularly predictive of adverse procedural outcome. This improved scheme may improve clinical decision making and provide a framework on which to base meaningful subgroup analysis in randomized trials assessing the efficacy of percutaneous coronary angioplasty.