The effects of age on the risks and results of heart valve replacement are often discussed in aortic stenosis, a degenerative condition of elderly patients. We assessed this factor in a study based on two groups of patients operated consecutively between 1968 and 1980. Group A comprised 127 patients of over 65 years of age (range 66 to 79 years, mean 70 +/- 3 years), and Group B: 259 patients of less than 65 years of age (21 to 65 years, mean 53 +/- 9 years). Preoperative assessment showed a higher proportion of women and significantly more severe lesions in Group A: a greater number of patients in functional class 3 or 4, cases of congestive cardiac failure, greater cardiac volumes, lower cardiac index. The incidence of coronary artery disease was also higher. A greater number of aortic valve bioprostheses was used in Group A. Operative mortality was significantly higher in Group A (15 p. 100, compared to 6 p. 100 in Group B: p less than 0.01). The influence of age was independent of other risk factors as shown by a multifactorial analysis taking several prognostic factors into account simultaneously. Long term survival was significantly lower in Group A, but their life expectancy did not differ from that of a control group. The causes of postoperative or late death were the same in both groups. At the end of the 6th year, 66 p. 100 of the survivors of surgery in Group A were in functional classes I or II compared to 77 p. 100 in Group B (NS). These results justify this type of surgery in elderly patients, even at advanced stages of their disease.(ABSTRACT TRUNCATED AT 250 WORDS)