The effects of acute reduction of left ventricular (LV) loading in valvular aortic stenosis (AS) were examined. Thirty-five consecutive patients with AS (peak-to-peak aortic valve gradient 66 +/- 26 mm Hg, aortic valve area 0.65 +/- 0.22 cm2) were given intravenous sodium nitroprusside (1 to 3-mu-g/kg/min) to reduce systolic aortic pressures by > 10 mm Hg (mean aortic pressure 99 +/- 15 to 80 +/- 15 mm Hg; p < 0.001). Overall, nitroprusside infusion resulted in little change in cardiac index (2.72 +/- 0.61 to 2.67 +/- 0.58 liters/min/m2; p = not significant). Individual patients had a range of responses. Fourteen patients (group 1) had an increase in cardiac index (2.42 +/- 0.59 to 2.74 +/- 0.67 liters/min/m2; p < 0.001), whereas 21 (group 2) had a decrease or no change (2.93 +/- 0.56 to 2.61 +/- 0.52 liters/min/m2; p < 0.001). Comparison of these subgroups showed that a cardiac index increase with nitroprusside was significantly predicted by a higher LV end-diastolic pressure (26 +/- 12 vs 15 +/- 6 mm Hg), lower LV ejection fraction (44 +/- 18 vs 62 +/- 12%), smaller aortic valve area (0.52 +/- 0.12 vs 0.74 +/- 0.22 cm2) and lower cardiac index (2.42 +/- 0.59 vs 2.93 +/- 0.56 liters/min/m2) (all values groups 1 and 2, respectively). It is concluded that there is a disparate response to acute vasodilation in AS. Potentially beneficial effects are seen in a subgroup of patients, especially those with increased filling pressures and impaired LV function. It is postulated that the hemodynamic effects of sodium nitroprus-side in patients with AS can be explained by the balance of effects on preload and afterload.