This study was done to assess the accuracy of various techniques of measuring the pressure gradient and valve area in patients with aortic stenosis (AS). In 18 patients with AS, the pressure gradient was quantitated from (1) simultaneous left ventricular and ascending aortic pressures (LV-AO), (2) nonsimultaneous LV-AO pullback, (3) LV and femoral arterial (FA) pressures unadjusted for the time delay of the FA tracing (LV-FA unadjusted), and (4) LV-FA adjusted for time delay. In comparison to simultaneous LV-AO, the pressure gradient was greater with LV-FA unadjusted and less with LV-FA adjusted for time delay (p less than 0.05). In nine patients with a mean gradient less than 35 mm Hg, the difference in valve area between simultaneous LV-AO and pullback averaged 0.17 +/- 0.10 cm2; between LV-AO and LV-FA unadjusted, the difference averaged 0.11 +/- 0.14 cm2; and between LV-AO and LV-FA adjusted, the difference averaged 0.52 +/- 0.36 cm2. These differences in valve area resulted in a discordant classification of the severity of AS in eight of the nine patients. Thus the use of an LV-AO pullback or an LV-FA gradient for assessing the severity of AS may yield inaccurate results, especially in patients with low (less than 35 mm Hg) gradients.