Aims/hypothesis Sulfonylureas (SUs) may impair outcome in patients with acute coronary syndrome. Most experimental studies of the myocardial effects of SU treatment are performed in non-diabetic models. We compared the effect of two widely used SUs, glibenclamide (gb) and gliclazide (gc), with high and low myocardial K ATP channel affinity, respectively, at therapeutic concentrations on infarct size, left ventricular (LV) function and myocardial glycogen, lactate and alanine content before and after ischaemia/ reperfusion (I/R). Methods Non-diabetic Wistar and diabetic Goto-Kakizaki rat hearts were investigated in a Langendorff preparation. Gb (0.1 μmol/l) and gc (1.0 μmol/l) were administrated throughout the study. Infarct size was evaluated after 120 min of reperfusion. Myocardial metabolite content was measured before and after ischaemia.Results Infarct size was smaller in diabetic hearts than in non-diabetic hearts (0.33±0.03 vs 0.51±0.05, p<0.05). Gb increased infarct size (0.54±0.04 vs 0.33±0.03, p<0.05) and reduced post-ischaemic LV developed pressure (60±3 vs 76±3 mmHg, p<0.05) and coronary flow (4.9±0.5 vs 7.1±0.4 ml min -1 g -1 , p<0.05) in gb-treated diabetic rats compared with untreated diabetic rats. On comparing gb-treated diabetic rats with untreated diabetic rats, glycogen content was reduced before (9.1 ±0.6 vs 13.6 ± 1.0 nmol/mg wet weight, p<0.01) and after ischaemia (0.9±0.2 vs 1.8±0.2 nmol/mg wet weight, p<0.05), and lactate (4.8±0.4 vs 3.2±0.3 nmol/mg wet weight, p<0.01) and alanine (1.38±0.12 vs 0.96±0.09 nmol/mg wet weight, p < 0.05) contents were increased during reperfusion. Gc-treatment of diabetic and non-diabetic rats did not affect any of the measured variables. Conclusions/interpretations Gb, but not gc, exacerbates I/R injury and deteriorates LV function in diabetic hearts. These effects of gb on diabetic hearts may be due to detrimental effects on myocardial carbohydrate metabolism.