Clinical and anatomic determinants of the initial success of percutaneous transluminal coronary angioplasty were prospectively evaluated in 826 patients enrolled in the Multi-Hospital Eastern Atlantic Restenosis Trial (M-HEART). The 639 men and 187 women ranged in age from 31 to 85 years. Successful angioplasty (residual stenosis less than 50% and no major complications) was achieved in 886 (88.6%) of 1,000 lesions. Success rates were uniform among the eight individual centers. Outcome was not influenced by gender, age or other clinical features, including severity and duration of angina, prior myocardial infarction, rest pain, transient ST segment elevation, history of smoking or diabetes. In contrast, procedural outcome was significantly associated with lesion-specific angiographic factors. Stenoses 60% to 74%, 75% to 89%, 90% to 99% and 100% were associated with success rates of 96%, 90%, 84% and 69%, respectively (p less than 0.001). Angioplasty was less successful in calcified than in noncalcified lesions (82% versus 90%, p less than 0.01), in thrombotic than in nonthrombotic lesions (82% versus 90%, p less than 0.05) and in lesions in the right coronary artery than in other vessels (84% versus 90%, p less than 0.01). Outcome was not related to other anatomic variables, including lesion location (proximal versus distal), vessel size, eccentricity, stenosis length or translesional gradient. By multivariate logistic regression, preangioplasty percent stenosis, right coronary artery location and lesion calcification were demonstrated to be significant independent predictors of angioplasty success. Alternative clinical and angiographic variables did not contribute to this regression model.(ABSTRACT TRUNCATED AT 250 WORDS)