Preliminary reports have documented the utility of balloon aortic valvuloplasty as a palliative treatment for high-risk patients with critical aortic stenosis, but the effect of this procedure on cardiac performance has not been studied in detail. Accordingly, 32 patients (mean age 79 years) with long-standing, calcific aortic stenosis were treated at the time of cardiac catheterization with balloon dilatation of the aortic valve, and serial changes in left ventricular and valvular function were followed before and after valvuloplasty by radionuclide ventriculography, determination of systolic time intervals, and Doppler echocardiography. Prevalvuloplasty examination revealed heavily calcified aortic valves in all patients, a mean peak-to-peak aortic valve gradient of 77 +/- 27 mm Hg, a mean Fick cardiac output of 4.6 +/- 1.4 liters/min, and a mean calculated aortic valve area of 0.6 +/- 0.2 cm2. Subsequent balloon dilatation with 12 to 23 mm valvuloplasty balloons resulted in a fall in aortic valve gradient to 39 +/- 15 mm Hg, an increase in cardiac output to 5.2 +/- 1.8 liters/min, and an increase in calculated aortic valve area to 0.9 +/- 0.3 cm2. Individual hemodynamic responses varied considerably, with some patients showing major increases in valve area, while others demonstrated only small increases. In no case was balloon dilatation accompanied by evidence of embolic phenomena. Supravalvular aortography obtained in 13 patients demonstrated no or a mild (less than or equal to 1+) increase in aortic insufficiency. Serial radionuclide ventriculography in patients with a depressed left ventricular ejection fraction (i.e., that less than or equal to 55%) revealed a small increase in ejection fraction from 40 +/- 13% to 46 +/- 12% (p less than .03). In addition, for the study group as a whole there was a decrease in left ventricular end-diastolic volume index (113 +/- 38 to 101 +/- 37 ml/m2, p less than .003), a fall in stroke-volume ratio (1.49 +/- 0.44 to 1.35 +/- 0.33, p less than .04), and no immediate change in left ventricular peak filling rate (2.05 +/- 0.77 to 2.21 +/- 0.65 end-diastolic counts/sec, p = NS). Serial M mode echocardiography and phonocardiography showed an increase in aortic valve excursion (0.5 +/- 0.2 to 0.8 +/- 0.2 cm, p less than .001), a decrease in time to one-half carotid upstroke (80 +/- 30 to 60 +/- 10 msec, p less than .001), and a small decrease in left ventricular ejection time (0.44 +/- 0.03 to 0.42 +/- 0.02 sec, p less than .001).(ABSTRACT TRUNCATED AT 400 WORDS)