Background. Allograft aortic valve replacement has gained widespread acceptance. However, there is little information about in vivo allograft valve function at rest and during exercise. Methods. Cardiac catheterization was performed to measure hemodynamic variables at rest and during supine bicycle exercise in 44 patients who had had aortic valve replacement using allograft valves or Bicer or St. Jude Medical prosthetic valves 19 to 27 mm in diameter. Sixteen patients received an allograft valve; 17, a Bicer valve; and 11, a St. Jude Medical valve. There were no significant differences between the three groups in age, body surface area, left ventricular end-systolic and end-diastolic volume indices, exercise cardiac index, exercise heart rate, or work load achieved. Left ventricular and ascending aortic pressures were measured simultaneously according to the transseptal method. Results. The mean pressure gradient was generally higher for the Bicer and St. Jude Medical valves than for the allograft valves, both at rest and during exercise. Significant differences were obtained in patients with small-sized valves (21 and 23 mm); pressure gradients were higher in the prosthetic valve groups. In patients with large-sized prosthetic valves (25 mm), there were no significant differences between the three groups at rest and during exercise. However, there was no pressure gradient at all for allograft valves. Conclusions. Exercise cardiac catheterization confirms that the allograft aortic valve is an ideal substitute from the hemodynamic aspect, particularly in patients with a small aortic root and in those who perform strenuous exercise.