To our knowledge, the study by Kawas et al. is the first observational study to describe a significant increase of dementia risk in older women who had been using hormones for five to nine years. The Leisure World Cohort Study was initiated in 1981 and consisted of women residing in a California retirement community (N = 8,801) who had completed a postal survey including details of estrogen therapy [ET]. After 22 years of follow-up, ET users (ever) had a 10% lower age-adjusted mortality than lifetime non-users. This risk was further decreased with increasing duration, but was not related to dose (0.625 mg vs. 1.25 mg). The relationship did not significantly change when adjusted for potential confounders, including exercise, body mass index, smoking, and medical histories related to mortality. The 90+ Study was initiated to directly examine surviving members of the cohort to determine clinical and functional status. Research participants who were alive and aged 90 and older on January 1, 2003 were invited to participate (N = 706 women, 90–106 years old). Prevalence of all-cause dementia was 45% in these women, but was not lower in ever-users of estrogen. Somewhat surprisingly, women who had used estrogen for five to nine years appeared to have an increased prevalence of dementia (odds ratio (OR) = 2.02, 95% CI 1.17–3.45), but longer or shorter duration of use was not associated with increased or decreased risk. This chapter discusses these results with regards to timing of exposure and age.