To assess the contribution of residual muscle perfusion in the infarcted territory to prevent ventricular remodeling, 24 patients with 1-vessel disease underwent coronary angiography and angioplasty of a critical left anterior descending coronary stenosis 18 ± 11 days after a first anterior myocardial infarction. The degree of stenosis was assessed using biplane quantitative angiography, whereas ventricular volumes, together with regional wall motion, were computed from single-plane ventriculography. Patients were reevaluated at 6 months after they had been subdivided according to the videointensity of the territory of the culprit vessel, as assessed from images obtained during main stem dye contrast injections before and immediately after angioplasty using a subtraction technique (group A, increased intensity [n = 15]; group B, no change [n = 9]), assuming that higher peak intensities reflect greater myocardial blood volume. There was a significant time · group interaction for ventricular volumes (diastolic, −13 ± 12% for group A vs +20 ± 24% for group B, p = 0.008; systolic, −15 ± 19% for group A vs +18 ± 36% for group B, p = 0.017), although no interaction was evident for the degree of resolution of coronary stenosis or the extent of recovery of regional dysfunction. The effects on volumes were paralleled by changes in ventricular end-diastolic pressure (−3 ± 7 mm Hg in group A vs +5 ± 6 mm Hg in group B, p = 0.006), although baseline clinical characteristics and medical regimen over the 6-month period were quite comparable between the 2 groups. In conclusion, despite late angioplasty of the culprit vessel, ventricular remodeling is prevented mainly when the procedure guarantees improved perfusion at the muscular level. The result is not necessarily mediated by recovery of regional systolic function.