The present study examined how Ca++ entry and intracellular Ca++ overload may contribute to the appearance of torsades de pointes in the setting of delayed repolarization. In anesthetized rabbits, the infusion of methoxamine and the selective I kappa s blocker almokalant (8.8 micrograms/kg.min) was associated with a lengthening of the QTU interval (37 +/- 2.6 ms, P < .001) and the appearance of torsades de pointes in 9/10 rabbits. In rabbits pretreated with nisoldipine (7.7 or 37 micrograms/kg i.v.), the incidence of almokalant-induced torsades de pointes was reduced to 7/10 (P = .5820 vs. vehicle) and 1/10 (P = .0006) rabbits, respectively. This occurred without attenuating the QTU-prolonging effect of almokalant (47 +/- 7.0 ms and 56 +/- 8.6 ms, respectively). Likewise, pretreatment with flunarizine (0.5 or 3.0 mg/kg i.v.) reduced the incidence to 1/6 (P = .0076) and 0/6 animals (P = .0009), respectively. In 10 of the rabbits that were given nisoldipine or flunarizine and did not experience torsades de pointes with almokalant, BAY K 8644 (0.11 mg/kg) was injected. In six of these rabbits, BAY K 8644 promptly induced torsades de pointes. In four vehicle-pretreated rabbits that experienced torsades de pointes with almokalant, acute injection of nisoldipine (37 micrograms/kg) abruptly suppressed the proarrhythmia. In separate experiments, rabbits were treated with ryanodine or BAPTA-AM and were subsequently administered almokalant. Compared with the vehicle-pretreated rabbits, these interventions did not significantly reduce the incidence of torsades de pointes (from 6/5 rabbits to 3/8 and 3/8 rabbits, respectively, P = .1776). The results demonstrate that Ca++ entry through the L-type Ca++ channel may be of crucial importance for the induction of torsades de pointes in the acquired long QT syndrome.