Objectives and Background. To assess the long-term results of percutaneous aortic valvuloplasty and aortic valve replacement in elderly persons, two similar nonrandomized series of patients greater-than-or-equal-to 75 years old treated hy one or the other method between January 1986 and March 1989 in the same institution were compared. Methods. Forty-six patients, 23 men and 23 women, with a mean age of 79.7 +/- 3.6 years (range 75 to 90) underwent percutaneous aortic valvuloplasty with use of the Cribier method (group 1). Twenty-three additional patients, 14 men and 9 women with a mean age of 78.4 +/- 2.4 years (range 75 to 86) underwent aortic valve replacement with a bioprosthesis (group 2). All of them suffered from severe calcified aortic stenosis. Clinical and hemodynamic status were similar in both groups. The mean follow-up period was 21.5 months (5 days to 60 months) in group 1 and 27.5 months (7 days to 61 months) in group 2. Results. Three patients (6.5%) in group 1 died within 5 days after percutaneous aortic valvuloplasty; 24 patients (52%) died during the follow-up period, 16 of whom died of recurrent cardiac failure. Of 16 patients (35%) subsequently operated on at an average of 15.8 months after percutaneous aortic valvuloplasty, 2 died at operation. Only three group 1 patients (6.5%) are still alive without subsequent aortic valve replacement. In group 2, two patients (8.7%) died postoperatively and three (13%) died during the follow-up period. All other patients (78%) are still alive and in New York Heart Association functional class I or II. The overall survival rate in group 1 was 75% at 1 year, 47% at 2 years and 33% at 5 years. In group 2, the survival rate was 83% at 1 and 2 years and 75% at 3 and 4 years. Conclusions. The results of percutaneous aortic valvuloplasty do not compare favorably with those of surgery in elderly people, and this treatment should not be recommended.