Reperfusion of the isolated rat heart following 10 min of coronary artery ligation under constant flow conditions results in the development of arrhythmias - premature ventricular contractions (PVCs), ventricular tachycardia (VT) and ventricular fibrillation (VF). Increasing concentrations of magnesium (0 - 4.8 mM) and/or potassium (2.5 - 10.0 mM) attenuate, while calcium (0.6 - 2.4 mM) exacerbates these arrhythmias. The protective effects of magnesium and potassium were additive. Magnesium reduced heart rate, perfusion pressure and developed tension. Potassium reduced perfusion pressure and increased developed tension. These haemodynamic effects contributed to the antiarrhythmic action of magnesium but did not completely account for its antiarrhythmic action. Calcium increased developed tension and heart rate and reduced perfusion pressure. Post-ligation administration of magnesium and potassium also protected against reperfusion arrhythmias. These results demonstrate that reperfusion arrhythmias are significantly affected by the ionic environments of the heart. Superoxide dismutase (5 - 20 Um1-1 ), glutathione (10 -5 - 10-3M), ascorbic acid (10-4 -5 x 10-4 M) and histidine (5 x 10-3 M) when given before coronary artery ligation attenuated the development of reperfusion arrhythmias. Mannitol (2 x 10-2 M) and catalase (100 - 300 Uml-1) did not have any significant effect on reperfusion arrhythmias when given alone but they did potentiate the effect of superoxide dismutase. Glutathione and a combination of superoxide dismutase, catalase and mannitol also reduced the incidence of reperfusion induced ventricular fibrillation when given just before reperfusion. Ferrous ion exacerbated the severity of reperfusion arrhythmias. Mannitol (2 x 10-2 M), catalase (100 uml-1) and histidine (5 x 10-3 M) when given before coronary ligation or just before reperfusion prevented the effect of ferrous ion while superoxide dismutase did not, indicating that the presence of ferrous ion is important for the production of hydroxyl radicals. Pretreatment with 6- OHDA attenuated the incidence of reperfusion arrhythmias but pre-ligation administration of allopurinol had no effect on reperfusion arrhythmias. By perfusing hearts with ferricytochrome C it was possible to show an increased reduction of ferricytochrome C during the first minute of reperfusion which could be prevented by the addition of superoxide dismutase and 6-OHDA treatment. These results provide evidence that oxygen free radicals are produced and may be important in the genesis of reperfusion induced arrhythmias in the isolated rat heart. Adenosine (10-6 M), verapamil (10 -8 - 10-7 M), ZK 36374 (10-10 - 10 -9 M) and sodium nitroprusside (10-6 -10-9 M) attenuate the incidence of reperfusion arrhythmias which may be via a coronary steal effect. Agents which affect arachidonic acid metabolism yielded conflicting results which may reflect nonspecific mechanisms other than inhibition of arachidonic acid metabolism. Glutathione and a mixture of superoxide dismutase, catalase and mannitol when given before coronary ligation and just before reperfusion reduced the increase in 86 rubidium efflux rate constant shown on reperfusion. The effect of glutathione on 86 rubidium efflux may be at least in part due to its vasodilator effect. Superoxide generation by xanthine/xanthine oxidase system increased the rate of efflux of 86 rubidium. A mixture of superoxide dismutase, catalase and mannitol also reduced the transient increase in the rate of release of 3H-noradrenaline shown to be produced on reperfusion after 10 min of ischaemia in the isolated rat heart.