Animal and human studies have indicated that the progression of atherosclerosis, notably in the coronary vessels, is accelerated by hypo-estrogenism and slowed down by estrogens. The principal favourable effect of endogenous or exogenous estrogens is not that they modify, quantitatively or qualitatively, blood cholesterol: it is that they prevent cholesterol from accumulating in the arterial wall. This favourable effect of estrogens is completed by other mechanisms which modify platelet aggregation, prostacyclin/thromboxane equilibrium, endothelin, etc., and which improve arterial wall morphology and vasoactive reactions. Some synthetic progestogens, given with oestrogens for endometrium control, may compromise these spectacular vascular benefits, but this is not the case with natural progesterone. Until now, the possible risk of thrombo-embolism has limited the prophylactic use of hormonotherapy and imposed contraindications. This risk is due to imbalance in hepatic (procoagulant effect) and peripheral (fibrinolytic effect) stimulations and therefore is to be feared specifically with oral oestrogens and their hepatic first-pass effect. In the absence of randomized trials, it is difficult to evaluate the true importance of this risk in the general population or in specific high-risk groups, but several surveys suggest that it persists even with the minimal doses currently administered orally. On the other hand, there is no known theoretical risk when oestradiol is given parenterally, which in France is the favoured method. This type of replacement therapy with oestradiol should be more often prescribed to subjects with atherogenic hyperlipidaemia, arterial hypertension or ischaemic coronary disease: not only it is not contraindicated, but it should soon become an indication.