When low doses of clofilium were administered to conscious rabbits, ventricular tachyarrhythmia (VT) with features typical of torsades de pointes developed. This arrhythmia was further studied and characterized in chloralose-anesthetized rabbits. In 20 of 20 rabbits, VT developed after a mean cumulative dose of 0.53 +/- 0.04 mumol/kg clofilium, provided an infusion of the alpha 1-agonist methoxamine (15 micrograms/kg/min) was given concomitantly. The arrhythmia was preceded by a marked prolongation of the QTU interval and the monophasic action potential duration, as well as signs of early afterdepolarizations (EADs). In seven of 10 rabbits receiving only clofilium (cumulative dose, 20.8 mumol/kg), no VT occurred (p less than 0.001 compared to the incidence in animals given both methoxamine and clofilium). In animals given both methoxamine and clofilium, pretreatment with prazosin (1 mg/kg i.v., n = 4) attenuated the arrhythmia, whereas diltiazem (0.5 mg/kg i.v., n = 4) or propranolol (0.5 mg/kg i.v., n = 8) was ineffective. Acute intervention with prazosin, isoproterenol, pinacidil, or magnesium sulfate promptly regularized the rhythm in animals with VT. Prazosin and pinacidil were equally effective in beta-blocked rabbits. When other agents known to retard the repolarization currents (sematilide, UK-68,798, LY97119, amperozide, and cesium chloride) were examined, a strong correlation (r = 0.99, p less than 0.001) between the potency of the drug to prolong the QTU interval and the proarrhythmic potential was obtained. This experimental model may represent an appropriate alternative for studying the acquired ("pause-dependent') long QT syndrome.