The Women's Health Initiative (WHI) hormone trials have been widely interpreted as demonstrating that combined menopausal hormone therapy (HT) fails to protect against-and may increase-cardiovascular disease (CVD), stroke, and dementia in menopausal women, regardless of whether initiated early in the menopause or later. This conclusion does not agree with results of large epidemiological studies showing protection by HT and by estrogen replacement alone (ET) against CVD and dementia. One possible reason for this inconsistency is that the epidemiologic data are confounded by "healthy user bias." Another possible explanation is that most women in the observational studies initiated ET or HT at or near the menopausal transition, at which point there is little or no arterial injury, whereas, in the WHI studies, older women, averaging approximately 12 years postmenopausal, many of whom would have had significant asymptomatic atherosclerosis, were treated. Substantial data demonstrate atheropreventive effects of estrogen before vascular damage occurs, whereas adverse effects of oral estrogen on thrombosis and inflammation may predominate once complex atheromas are present. Similarly, the excess of dementia observed in older WHI women treated with oral conjugated estrogen could be due to cerebral thromboses (multi-infarct dementia). Given the uncertain relevance of the WHI (and other published randomized clinical trials) to initiation of HT in perimenopausal women, and its subsequent continuation for atheroprevention, new trials will be needed to resolve whether early intervention with estrogen may prevent CVD and/or dementia. The Kronos Early Estrogen Prevention Study (KEEPS), which began in mid-2005, is a randomized, controlled multicenter trial of HT in recently menopausal women. It will examine surrogate end points as well as risk factors for atherosclerosis.