The mechanisms for the different proarrhythmic potential of antiarrhythmic drugs in the presence of comparable QT prolongation are not completely understood. The reasons for the lower proarrhythmic potential of amiodarone as compared with other class-III antiarrhythmic drugs such as sotalol, a fact that has been well established for years, is insufficiently known. Therefore, the aim of our study was to assess the different electrophysiologic effects of amiodarone and sotalol in a previously developed experimental model of proarrhythmia. In eight male rabbits, amiodarone (280-340 mg/d) was fed over a period of six weeks. Hearts were excised and retrogradely perfused. Up to eight simultaneous epi- and endocardial monophasic action potentials (MAP) were recorded. Results were compared with sotalol-treated (10-50-100 microM) hearts (n = 13). Amiodarone and sotalol (50 microM and 100 microM) led to a significant increase in QT interval (mean increase: amiodarone: 31 +/- 6 ms; sotalol: 41 +/- 4 ms and 61 +/- 9 ms) and MAP-duration (mean increase-MAP90: amiodarone: 20 +/- 5 ms; sotalol: 17 +/- 5 ms and 25 +/- 8 ms) (P < 0.01). In bradycardic (AV-blocked) hearts, MAP-recordings demonstrated reverse-use dependence and a significant increase in dispersion of repolarization (MAP90) in the presence of sotalol (P < 0.01), but not in amiodarone-treated hearts (10%; p = ns). Sotalol led to early afterdepolarizations (EAD) and torsade de pointes (TdP) after lowering of potassium concentration (6 of 13 hearts). In amiodarone-treated, hypokalemic hearts, no EAD or TdP occurred. Sotalol changed the MAP configuration to a triangular pattern (ratio-MAP90/50: 1.52 as compared with 1.36 at baseline) whereas amiodarone caused a rectangular pattern of MAP prolongation (ratio-MAP90/50: 1.36). In conclusion, these results show no direct correlation between the occurrence of TdP and the degree of QT prolongation. Several factors including reverse-use dependence, dispersion of repolarization, and the propensity to induce early afterdepolarizations but also differences in the action potential configuration may help to understand proarrhythmic side effects of drugs.