Contrary to widespread irrational belief, estrogen replacement therapy (ERT) bears very little risk and, on the contrary, exerts a protective effect against atherosclerosis and cardiovascular disease. This protection most likely results from one or both of the following mechanisms: (1) a drive of the lipidic metabolism towards a “favorable” profile (decrease of total- and LDL-cholesterol and increase in HDL-cholesterol levels); (2) a favorable prostacycline/thromboxane balance towards vasodilatation and anti-aggregation. This is the case when natural estrogens are used, since synthetic estrogens (namely ethinylestradiol) overstimulate the hepatic production of proteins, such as renin substrate and angiotensinogen, with an increased risk of hypertension, vasoconstriction and platelet aggregation. Conjugated equine estrogens are not truly physiologic for human use and some of their estrogenic components can also predispose to hypertension and accumulate. Non-orally administered natural estrogens have less — almost no — metabolic impact and this could hypothetically result in an impaired protective effect against cardiovascular disease; the presently available estrogen implants are not satisfactory and their use frequently leads to estrogen accumulation. Hypertension and diabetes are no longer considered to be contra-indications to ERT: natural estrogens can even improve hypertension as well as glucose tolerance. Estrogen use, at least by the oral route, slightly increases the incidence of gallstones. The sole other well-documented risk of unopposed (i.e., without administration of any progestogen) ERT resides in a 2x–3x greater risk of developing a carcinoma of the endometrium, the effect being dose- and duration-dependent. A progestogen challenge test is mandatory before initiating ERT and the addition of a progestogen must be considered in case of long-term ERT. Furthermore, patients having received ERT remain at increased risk for endometrial carcinoma development for many years and thus require careful post-treatment surveillance. As far as breast cancer is concerned, and despite the clear involvement of exogenous estrogens in its etiology, the bulk of the data offers reassurance that low-dose exogenous estrogens will not dramatically increase its incidence. However, it remains controversial whether or not high doses and long duration could still, however, slightly increase the risk. Therefore, regular breast screening (X-ray) is mandatory and it is recommended to use the lowest estrogen dose possible that can treat menopausal complaints and/or prevent osteoporosis. In this respect, estrogen doses could perhaps still be reduced (to the equivalent of 0.3 mg conjugated equine estrogens), provided additional measures such as calcium supplementation and physical exercise are taken. Synthetic progestogens can impair carbohydrate metabolism but this is seldom clinically relevant. On the contrary, in relation to their degree of androgenicity and to the dose used, they induce a potentially harmful lipidic profile towards promotion of atherosclerosis and increased risk of arterial vascular disease. When added sequentially or continuously to ERT, androgenic progestogens do counteract the supposedly favorable lipidic profile resulting from ERT and the net result depends on the doses of both steroids, on their route of administration, as well as on the nature of the progestogen. In this respect, it might be recommended to preferentially use progestogens with little or no androgenicity, such as micronized progesterone, dydrogesterone and possibly medrogestone; low doses of cyproterone acetate can be considered and the marketing of progestogens of the “third” generation (such as desogestrel, that does not counteract estrogens as far as the lipidic profile is concerned) is awaited. It should be kept in mind that synthetic progestogens could also facilitate the development of hypertension (while natural progesterone has been reported to exhibit hypotensive effects) and hypothetically disturb the prostacyclin/thromboxane balance towards an increased cardiovascular risk. However, the demonstration that addition of a progestogen might partially suppress the cardioprotective effect of ERT awaits epidemiological evidence. The addition of a progestogen to ERT is thus justified only in non-hysterectomized women. Their solely demonstrated indication is indeed the prevention of the development of endometrial carcinoma. This requires a 10–12-day duration of administration of an adequate progestogen at a sufficient dose each “cycle” or month. Progestogen addition will, unfortunately, induce more or less tolerable side-effects and result in bleeding, that may or may not be accepted by the patient. Continuous combined estro-progestative treatment represents an interesting regimen for 60% of the patients who, fortunately, become and remain amenorrheic, exhibiting an atrophic endometrium.