1. We examined whether N-hydroxyethyl-1-deoxynojirimycin (miglitol), a new human anti-diabetic drug with effects to inhibit alpha-1, 6-glucosidase glycogen debranching enzyme and reduce the glycogenolytic rate as well as to inhibit alpha-1,4-glucosidase, could reduce infarct size in the rabbit heart. Rabbits were subjected to 30-min coronary occlusion followed by 48-h reperfusion. 2. The infarct size as a percentage of area at risk was not reduced by pre-ischaemic treatment with 1 mg kg(-1) miglitol (42.7+/-4.0%, n=10) compared with the saline control group (41.7+/-2.3%, n=10). However, it was significantly and dose-dependently reduced by pre-ischaemic treatment with 5 or 10 mg kg(-1) of miglitol (25.7+/-4. 5%, n=10, and 14.6+/-2.4%, n=10, respectively) without altering the blood pressure, heart rate or blood glucose level. However, there was no evidence of an infarct-size reducing effect after pre-reperfusion treatment with 10 mg kg(-1) of miglitol (35.0+/-3.0%, n=10). 3. Another 40 rabbits given 1, 5 and 10 mg kg(-1) of miglitol or saline before ischaemia (n=10 in each) were sacrificed at 30 min of ischaemia for biochemical analysis. Miglitol preserved significantly the glycogen content, and attenuated significantly the lactate accumulation in a dose dependent manner in the ischaemic region at 30 min of ischaemia. 4. Pre-ischaemic treatment, but not pre-reperfusion treatment, with miglitol markedly reduced the myocardial infarct size, independently of blood pressure and heart rate. A dose-dependent effect of miglitol on infarct size, glycogenolysis and lactate formation suggests that the mechanism may be related to the inhibition of glycogenolysis. Thus, miglitol may be beneficial for coronary heart disease as well as diabetes mellitus.