When aortic valve replacement is performed in a patient with a small anulus, significant obstruction of the left ventricular outflow tract may remain. Most prostheses are obstructive in the smaller sizes, and enlargement of the aortic anulus may be required to allow placement of a larger valve. To evaluate the hemodynamic performance of two commonly used tissue prostheses, the Ionescu-Shiley pericardial and Carpentier-Edwards porcine valves, 22 patients with either the 19 or 21 mm size were electively studied at rest and after exercise at a mean of 15 months after operation. The resting mean transvalvular gradient for 19 mm Ionescu-Shiley pericardial valves (n = 7), 10.6 +/- 9.2 mm Hg, was significantly lower than that for 19 mm Carpentier-Edwards valves (n = 3), 33.3 +/- 2.1 mm Hg, p less than 0.01. Following exercise, the mean gradient for 19 mm Ionescu-Shiley pericardial valves rose only to 13.8 +/- 8.5 mm Hg. No exercise data were available for the 19 mm Carpentier-Edwards valve. Among patients with 21 mm Ionescu-Shiley pericardial valves (n = 7), the mean transvalvular gradient at rest was 5.6 +/- 9.5 mm Hg, not significantly different from that of patients with 21 mm Carpentier-Edwards valves (n = 5), 9.8 +/- 18.3 mm Hg. After exercise, the gradients rose to 16.0 +/- 10.0 mm Hg and 25.5 +/- 23.8 mm Hg for the Ionescu-Shiley pericardial and Carpentier-Edwards valves, respectively (no statistical significance). Cardiac index was not different between groups. Gradients were not significantly higher in patients with body surface areas greater than 1.5 m2. It is concluded that the 19 and 21 mm Ionescu-Shiley pericardial valves possess excellent hemodynamics, even after exercise. This valve appears hemodynamically superior to the Carpentier-Edwards valve, particularly in the 19 mm size. Procedures to enlarge the aortic anulus are usually unnecessary when small Ionescu-Shiley pericardial valves are used, even in patients who have large body surface areas.