Reports on effects of ATP-dependent K+ channel modulating drugs on ischaemia-induced cardiac arrhythmias have been scarce and contradictory. The channel blocking agent glibenclamide (glyburide) has been considered as an antiarrhythmic candidate, because it antagonizes the ischaemic K+ efflux and the shortening of the refractory period. In the present investigation its effects were tested, therefore, in rat hearts with coronary occlusion and reperfusion. In untreated hearts, tachyarrhythmias occurred during the reperfusion, and less pronounced during the coronary occlusion itself. Large amounts of adenosine and its degradation products were released during the coronary reperfusion, particularly from hearts which developed ventricular fibrillation. Glibenclamide (0.1 and 1.0 micromol/l perfusion fluid) neither antagonized the ischaemic nor the reperfusion arrhythmias. Ischaemic arrhythmias were even intensified. Also in control hearts without coronary occlusion, pro-arrhythmic effects of glibenclamide were observed. Furthermore, the coronary flow was considerably decreased by the drug, and the release of adenosine and its metabolites was significantly increased. Sodium nitroprusside antagonized the glibenclamide-induced decrease in the coronary flow, but did not prevent the arrhythmias. The Ca2+ channel blocking agent gallopamil increased the coronary flow, decreased the adenosine release, and antagonized the arrhythmias in hearts with and without glibenclamide. In conclusion, the present findings do not favour the idea of an antiarrhythmic effect of glibenclamide. Rather, some propensity to the occurrence of arrhythmias can be produced by the drug.