The supposed direct arrhythmogenic property of endothelin-1 (ET-1) has not yet been clearly proven. Our study aimed to characterize the electrophysiological changes during left anterior descending artery (LAD) occlusion and intracoronary (i.c.) ET-1 infusion, and to differentiate between the supposed direct and ischemic arrhythmogenic actions of ET-1 in a canine model. Changes of monophasic action potential duration (MAPD90) and upstroke velocity (UV) are capable of detecting local ischemic changes. Left and right ventricular endo- (LVEND, RVEND) and epicardial (LVEP, RVEP) monophasic action potentials were recorded. MAPD90, monophasic action potential dispersion (MAPDISP) and UV were determined. In group A (n = 8) 30 min LAD occlusion was followed by a 60 min reperfusion period. In groups B and C ET-1 was administered into the LAD at rates of 30 (n = 8) and 60 pmol/min (n = 10), respectively. In group A after the LAD occlusion both MAPD90 and UV decreased significantly in the LAD region (LVEP and LVEND 18 +/- 3% and 10 +/- 1%, p < 0.05, and 65 +/- 4% and 52 +/- 8%, respectively, p < 0.05; control and 30 min values in all groups), whereas the increase in MAPDISP remained unchanged. No severe arrhythmias were noticed in this group. In group B, both MAPD90 and MAPDISP increased significantly (LVEP and LVEND 11 +/- 4% and 18 +/- 3%, p < 0.05; MAPDISP 200 +/- 40%, p < 0.05), whereas UV remained unchanged at the end of the infusion. Early afterdepolarizations (EADs) were present in three instances. In group C both MAPD90 and MAPDISP increased significantly (LVEP and LVEND 12 +/- 5% and 26 +/- 8%, respectively, p < 0.05; MAPDISP 215 +/- 30%, p < 0.05) and UV decreased slightly in the LAD region. EADs were observed in five instances. Severe arrhythmias were observed in both groups B and C. We concluded that MAP prolongation, increased MAP dispersion and development of EADs all contribute to the arrhythmogenic action of ET-1. The lack of the almost prompt decrease of UV and MAPD90 which was observed in group A in groups B and C strongly supports the probability of a direct arrhythmogenic effect of ET-1.