A retrospective analysis of an institutional experience with aortic valve replacement (AVR) in patients 70 years of age or older during 1976 to 1987 was performed. The study was prompted in part by the current interest in palliative aortic valvoplasty, an interest based to a certain extent on the impression that AVR in the elderly has a high mortality. The mean age of the patients was 75.0 +/- 4.0 years (+/- the standard deviation) (range, 70 to 89 years). Eighty-three percent of patients received porcine valves and 17%, mechanical valves. Preoperatively 32% were in New York Heart Association class III, and 59% were in class IV. Operative mortality was 5.6% for elective isolated AVR for aortic stenosis (19% of all patients), 8.2% for all isolated AVR (38%), and 12.4% overall. Concomitant operative procedures were done in 62.0%; AVR with coronary artery bypass grafting (42%) had an operative mortality of 14.3%. Multivariate analysis showed significant predictors of operative mortality to be emergency operation (p less than 0.01), isolated aortic regurgitation (p = 0.01), and previous cardiac operation (p = 0.02). Follow-up (34 +/- 27 months) was 94% complete. Five-year survival from late cardiac-related death was 81.0%. The constant yearly hazard rate for late death for patients 70 years of age or older who underwent AVR was 5.42% per year, which is similar to the 5.77% per year rate calculated for age-matched and sex-matched controls. Five-year freedom from reoperation was 99%; from late thromboembolic complications, 91%; and from late anticoagulant-related complications, 94%. Freedom from all valve-related morbidity and mortality was 61% at 5 years.(ABSTRACT TRUNCATED AT 250 WORDS)