BACKGROUND Independently of the problem of atherogenesis, the amount and type of fat intake influences the risk of cardiac arrhythmias. However, the relative effectiveness of different fats and the underlying mechanisms are controversial. The aim of the present study was to compare the effects of various oil-enriched diets on the risk of ventricular arrhythmias in rat hearts under conditions of ischaemia and reperfusion and to help clarify the mechanisms underlying the differing effects of the oils on the occurrence of arrhythmias. METHODS Over a 10-week period, we studied five groups of young male Wistar rats given a low-fat chow diet or one enriched with 10% hydrogenated coconut oil, corn oil, linseed oil or sardine oil. Electrocardiograms were recorded from the isolated hearts (Langendorff preparation) perfused with a modified Krebs-Henseleit solution. Ischaemia was induced by a 20 min occlusion of the left anterior descending coronary artery. In another series of experiments, a 10 min occlusion was followed by a 20 min reperfusion period. The times between the first occurrence of extrasystole and the incidence of ventricular tachycardia and fibrillation were determined. The size of the ischaemic zone was assessed using malachite green. The fatty acid composition of the myocardial tissue was analysed using gas chromatography. RESULTS An increase in the risk of ventricular arrhythmias under conditions of both ischaemia and reperfusion was obvious in the rats that consumed large quantities of saturated fatty acids (coconut oil) and in the group with a very low intake of fat. Polyunsaturated fatty acids (PUFAs), particularly fish oil, exerted a protective effect. The incidence of ventricular fibrillation was 75% in the low-fat group, 67% in the coconut-oil group, 44% in the corn-oil group, 40% in the linseed-oil group and 10% in the fish-oil group. The time until the first occurrence of extrasystole, the incidence of ventricular tachycardia and the incidence of reperfusion-induced ventricular fibrillation were influenced in a similar manner. The size of the ischaemic zone was significantly reduced in the groups given diets enriched with PUFAs. All protective effects were abolished, however, by cyclooxygenase inhibition with aspirin. The fatty acid composition of myocardial tissue, the ratio of n-3 to n-6 fatty acids and the double-bond index were significantly affected by the various diets. CONCLUSIONS Whereas saturated fatty acids are obviously proarrhythmic, diets enriched with n-6 or n-3 PUFAs both exert antiarrhythmic effects. Although n-3 fatty acids seem to be more effective, cardioprotection cannot simply be related to the replacement of n-6 by n-3 fatty acids in cardiac membrane phospholipids, given the beneficial effects of corn oil. In any case, replacement of n-3 by n-6 fatty acids is not the underlying mechanism. The overall reduction of prostaglandin formation cannot be the primary mechanism because the beneficial effects of diets rich in PUFAs were abolished by cyclooxygenase inhibition. We conlcude that, besides prostacyclin (PGI2 or PGI3), membrane fluidity and accompanying alterations in functional membrane proteins (e.g. protection from calcium overload) are key factors apart from vascular effects that influence the size of the ischaemic zone.