We investigated aortic valve hemodynamic performance and perioperative left ventricular function in 50 patients (mean [+/- SD] age, 64 +/- 9 years; 34 men, 16 women) undergoing elective aortic valve replacement, using an aortic homograft (n = 20), a Toronto stentless porcine valve (n = 20), or a stented bioprosthesis (n = 10), by transesophageal echocardiography combined with high-fidelity cavity pressure recordings and thermodilution cardiac output measurements. Thirty-nine patients had aortic stenosis; 11 had predominant regurgitation. Thirteen patients with concomitant coronary artery stenosis underwent grafting. Left ventricular mass index in all patients was 280 +/- 110 g/m2. The transvalvular pressure drop and energy consumption were significantly higher with stented than stentless valves (5 with aortic homograft and 11 with Toronto valve, with matched age and valve size; 20 +/- 12 versus 3 +/- 9 mm Hg; 21% +/- 13% versus 8% +/- 8%, both p < 0.01). However, there was no difference in these variables between the Toronto valve and the aortic homograft (3 +/- 12 versus 2 +/- 10 mm Hg; 5% +/- 14% versus 2% +/- 12%, both p > 0.05), although the Toronto valves (normalized to body surface area) were larger than the aortic homografts (14.4 +/- 1.9 versus 12.6 +/- 1.8 mm/m2, p < 0.01). There was no significant difference in left ventricular stroke volume index or stroke work index in the systemic circulation, either between stentless and stented valves or between aortic homografts and Toronto valves, although the cross-clamp time required to insert a stentless valve was 20 minutes longer than that for a stented valve.(ABSTRACT TRUNCATED AT 250 WORDS)