Epidemiological and interventional studies indicate that dietary n-3 PUFA reduces mortality due to coronary heart disease (CHD). They act at a low dose, since one or two meals with fatty fish per week is sufficient to provide protection when compared with no fish intake. These fatty acids are effective in providing primary prevention in low- and high-risk subjects and secondary prevention. At high doses, dietary n-3 PUFAs have several beneficial properties. First, they act favourably on blood characteristics: they are hypocholesterolemic and hypotriglyceridemic; they reduce platelet aggregation; they exhibit antithrombotic and fibrinolytic activities; they reduce blood viscosity and they exhibit antiinflammatory action. Second, they reduce ischemia/reperfusion-induced cellular damage. This effect is apparently due to the incorporation of eicosapentaenoic acid in membrane phospholipids. Third, they reduce ischemia and reperfusion arrhythmias. All the effects exerted by n-3 PUFAs at high doses are incompatible with the beneficial action on CHD mortality in humans observed at low doses, where their main properties are related to circulation in the form of free fatty acids. Numerous experimental studies have indicated that low concentrations of exogenous n-3 PUFAs reduce the severity of cardiac arrhythmias. This effect is probably responsible for the protective action of n-3 PUFA on CHD mortality. Further studies are necessary to confirm this assumption in animals. Such studies should take account of the fact that only a low dose of n-3 PUFA (20 mg/kg/day) is necessary to afford protection. Furthermore, since the beneficial effect of n-3 PUFAs on CHD mortality is observed in fish eaters versus no-fish eaters, and since populations in industrialised countries consume excess n-6 PUFAs, control animals in long-term dietary experiments should be fed a diet with only n-6 fatty acids as a source of PUFAs.