The purpose of this study were to determine the specificity of the head-up tilt test in normal subjects when a graded isoproterenol infusion is used, and to evaluate the role of dynamic ventricular volume change during head-up tilt as a mechanism of syncope. We prospectively studied 12 normal volunteers, each of whom underwent an upright tilt test for 10 minutes at 80 degrees with and without an infusion of isoproterenol. A subgroup of five subjects had a third tilt test during administration of a combination of esmolol and isoproterenol. Blood pressure, heart rate, and left ventricular volumes and flow (obtained with Doppler echocardiography) were recorded in the following sequence: while supine, during upright tilt, while supine with isoproterenol, and during upright tilt with isoproterenol. During the initial head-up tilt, one subject had syncope. An additional eight subjects had presyncope or syncope during head-up tilt with isoproterenol. The remaining three subjects were asymptomatic. In subjects with syncope or near-syncope ("responders"), heart rate increased with isoproterenol but decreased markedly, to 76 +/- 5 beats/min, by the end of the protocol. Systolic blood pressure rose slightly above baseline during isoproterenol but fell from 118 +/- 4 to 85 +/- 5 mm Hg during head-up tilt with isoproterenol. The three asymptomatic subjects had only one significant change, an increase in heart rate with isoproterenol. In the five responders undergoing three tilt tests, left ventricular volume decreased significantly at end diastole (94 +/- 25 vs 58 +/- 22 ml) and end systole (34 +/- 13 vs 18 +/- 6 ml) when supine baseline is compared with initial upright tilt.(ABSTRACT TRUNCATED AT 250 WORDS)