The continuity equation is valid for a population of aortic valves, but its accuracy in individual valves is uncertain. In bileaflet prostheses, obstruction to forward flow is small and individual variability in opening behaviour is minimal. In these valves, the area of the effective orifice should be close to that measured by the manufacturer. A total of 57 patients aged 58 +/- 11 years were studied at a mean of 3.6 months after implantation with a CarboMedics aortic prosthesis. Nine had additional implants in the mitral position and all prostheses were clinically normal. Peak subaortic and transaortic velocities were averaged over 5 beats. The area of the effective orifice was significantly different between the four diameters (P < 0.00001), and the correlation between the effective and measured area of the orifice was moderate (r(s) = 0.73, P < 0.00003). The 95% range for the differences between individual pairs of values, however, was 0.16 +/- 0.61 cm2. Discrepancies probably arose in the estimation of subaortic cross-sectional area and subaortic velocity. Thus, the continuity equation may be inaccurate in an individual prosthetic valve when functioning normally.