Estrogen has the potential to influence brain processes implicated in Alzheimer's disease pathogenesis. With the loss of ovarian estrogen production after menopause, estrogen-containing hormone therapy might be expected to influence the risk of Alzheimer's disease. Observational data link use of hormone therapy to reductions in Alzheimer risk, but experimental evidence from the Women's Health Initiative Memory Study trial demonstrates that oral estrogen, with or without a progestin, increases the incidence of dementia for postmenopausal women age 65 years or older. Mechanisms of harm in this setting are unknown. Bias and unrecognized confounding in observational research are leading candidates for discrepant results between observational studies and the Women's Health Initiative Memory Study trial. Studies are also distinguished by differences in outcome measures, hormone therapy formulations, prevalence of menopausal symptoms among study participants, and participant age. Finally, Women's Health Initiative Memory Study findings may not generalize to estrogen use by relatively young women during the menopausal transition or early postmenopause, a class of women who were ineligible for the Women's Health Initiative Memory Study trial. In observational studies, hormone therapy exposure often included use by younger women for menopausal vasomotor symptoms. Although there is no clinical trial evidence that hormone therapy at any age protects against Alzheimer's disease, it remains to be determined whether the age at which hormone exposure occurs or the timing of hormone therapy initiation in relation to the menopause (the critical window hypothesis) modifies treatment outcomes on dementia risk.