Neurally mediated hypotension and bradycardia are believed to be common causes of syncope. We used the "upright-tilt test" (duration, less than or equal to 10 minutes) with or without an infusion of exogenous catecholamine (isoproterenol [1 to 5 micrograms per minute], given intravenously) to elicit bradycardia, hypotension, or both in 24 patients with recurrent syncope and in 18 control subjects. A conventional electrophysiologic test performed before the tilt test was positive in 9 of the 24 patients, revealing arrhythmias that may have caused recurrent syncope, but was negative and thus nondiagnostic in 15 patients. The tilt test alone (i.e., without isoproterenol) induced symptomatic bradycardia or hypotension in 1 of the 9 patients with positive electrophysiologic tests (11 percent), 4 of the 15 patients with negative electrophysiologic tests (27 percent), and none of the controls. When the isoproterenol infusion was administered during the tilt test, 9 of the 11 patients with negative electrophysiologic and tilt tests had syncope, marked slowing of the heart rate, and hypotension. In contrast, isoproterenol was associated with tachycardia and only a slight decline in arterial pressure in the 8 remaining patients with positive electrophysiologic tests and the 18 control subjects, and syncope developed in only 1 of the 8 patients with positive electrophysiologic tests and negative tilt tests (13 percent) and 2 of the 18 control subjects (11 percent). We conclude that an isoproterenol infusion administered in conjunction with the upright-tilt test may be useful for identifying susceptibility to neurally mediated recurrent syncope.