Seventeen patients with recurrent paroxysmal supraventricular tachycardia (SVT) underwent serial electrophysiologic studies to compare the effects of i.v. sotalol (1.5 mg/kg) and i.v. metoprolol (0.15 mg/kg). The plasma concentrations of sotalol (2.1 +/- 1.1 microgram/ml) and metoprolol (67 +/- 15 ng/ml) were within the therapeutic range. Before drug administration, sustained SVT could be reproducibly induced in all patients. Sotalol prevented induction of sustained SVT in 10 of 17 patients (59%) and metoprolol in 4 (28%) (p less than 0.05). In 6 of 8 patients with atrioventricular (AV) nodal reentrance, the site of action of sotalol was the anterograde or the retrograde limb, reflecting an increase in refractoriness in both pathways of the circus movement. In 4 of 9 patients with AV reentrance, the site of action of sotalol was exclusively the AV nodal pathway; conduction through the extranodal accessory tract appeared to be unchanged, but its anterograde effective refractory period was prolonged (from 285 +/- 25 to 322 +/- 28 ms, p less than 0.001; mean +/- standard deviation). In the 7 patients in whom sotalol did not prevent sustained SVT, the tachycardia cycle length increased from 347 +/- 42 to 392 +/- 45 ms (p less than 0.01). Compared with sotalol, metoprolol had qualitatively similar but quantitatively less potent effects on the AV nodal pathways; however, different from sotalol, metoprolol had no effect on extranodal accessory tracts. The study suggests that at therapeutic plasma concentrations, sotalol would be effective in preventing clinical SVT in a significant proportion of patients refractory to metoprolol; because sotalol not only has beta-blocking properties but also results in acute prolongation of the action potential duration, this combination of class II and III activity may contribute to its superior prophylactic efficacy compared with pure beta blockade.