Over a 4-year period, 120 transluminal coronary angioplasties were performed in 102 patients older than 75 years. There were 56 men and 46 women aged from 75 to 89 years (mean: 78 +/- 3 years) presenting with the following characteristics: left ventricular ejection fraction 60 +/- 11 percent; severe angina (class III or IV) 86 percent; history of myocardial infarction 43 percent; one-vessel lesion 39 percent; 66 percent of the arteries dilated were calcified. Altogether, 158 vessels (1.3 per procedure) were dilated: 1 vessel in 74 percent of the patients, 2 in 20 percent and 3 in 6 percent. The primary success rate was 80 percent per lesion (126/158 lesions) and 77 percent per procedure (92/120 procedures). This primary success rate was significantly higher when the stenosis dilated was not calcified (88 versus 75 percent, P < 0.05) and in cases with stenosis rather than complete occlusion (83 versus 53 percent, P < 0.01). Three patients died (3 percent) and 9 (7.5 percent) developed infarction with Q wave, but no emergency bypass was needed. The first 79 patients could be followed up for a mean period of 23 +/- 13 months (range: 8-61 months), and no patient was lost sight of. During that period, 11 patients died (including 7 of cardiac cause), 2 had a non-lethal infarction, 7 underwent distant aortocoronary bypass and 18 had a second angioplasty for restenosis. The long-term survival rate (Kaplan-Meier) was 83 +/- 6 percent at 4 years (90 percent when the angioplasty was successful and 73 percent when it failed; P < 0.02). Among the 65 survivors, 73 percent no longer had angina, 96 percent are in the New York Heart Association class I or II, and 92 percent are still on oral anti-angina therapy. Thus, transluminal coronary angioplasty can be performed in very old patients with good efficacy and an acceptable complication rate. The improvement obtained persists for a long time when the angioplasty is successful.