Background. There is increasing use of hormone replacement therapy (HRT) by post-menopausal women. Observational epidemiological studies have shown reductions in cardiovascular risk factors in HRT users in the USA, but no randomized controlled trials of HRT have been carried out in the primary practice setting. Previous studies of cardiovascular risk factors have shown a variety of responses according to type of progestagen and oral or topical administration. None has examined the effect of route using an identical progestagen.Objectives. Our aim was to establish differences, if any, in alteration in cardiovascular risk factors with HRT in post-menopausal women according to route of administration of HRT, oral, transdermal and implant, using first oestrogen alone then oestrogen plus norethisterone, or testosterone for implant.Methods. Subjects were recruited by letter of invitation to women aged 50–65 years from lists in general practices local to the Charing Cross Hospital Lipid Clinic in West London. Their menopausal status was confirmed and they were randomized to one of three treatment groups or acted as controls. They attended for three visits; at baseline, HRT was initiated as oestrogen alone, oral or transdermal. At the 3-month visit, HRT with the progestagen, norethisterone, was given cyclically, continuously or transdermally until the final visit at 6 months. A separate group of women from the menopause clinic at Chelsea and Westminster Hospital were studied on oestrogen implant then on implanted oestrogen and testosterone. The outcome measures studied were the separate effects of the four regimes as compared with controls on lipoproteins, glucose, insulin, fibrinogen, factor VII and E-selectin, together with weight, waist:hip ratio and blood pressure.Results. The continuous combined oestrogen–progestagen therapy had similar effects on cardiovascular risk factors as oestrogen with cyclical progestagen. All regimes lowered low-density lipoprotein cholesterol, the oral route being more potent than the parenteral; the effect of transdermal HRT was similar to the implant. Lp(a) was reduced only with the oral route. Reductions in factor VII and E-selectin were observed in both the oral and transdermal routes. There was no increase in body mass index, waist:hip ratio, blood pressure or glucose and insulin levels with any of the HRT regimes used. Systolic blood pressure was reduced with the transdermal route.Conclusions. This study supports the evidence that oestrogen–progestagen HRT, both oral and transdermal, although attenuating some of the benefit of oestrogen alone on fibrinogen and high-density lipoprotein, significantly reduces cardiovascular risk factors, which should diminish post-menopausal risk of coronary disease.