To investigate the relationship between hypertension and Alzheimer's disease(AD) and the change of Alzheimer's patients' blood pressure(BP) before and after the onset of AD, we conducted this epidemiological study. Subjects for this study were individuals who participated in a large scale, randomized controlled trial of nutritional intervention from 1984 to 1991. Participants were initially screened for dementia using Chinese Mini-Mental State Examination (CMMS) and Activities of Daily Living (ADL). Positive subjects were subsequently administered a detailed neuropsychological and neurobehavioral examination. The diagnosis of AD was made by a consensus conference of psychiatrists using Diagnostic And Statistical Manual Of Mental Disorders-Fourth Edition(DSM-IV) criteria. 16488 subjects were examined and 301 were diagnosed as AD. We compared the prevalence of AD in different populations that were stratified with 1984's systolic or diastolic blood pressure(those four stratifications being high blood pressure, borderline blood pressure, normal, low blood pressure), and compared the change of blood pressure of 301 AD patients between 1984 and 1999-2000, which is before and after the onset of AD respectively. Multiple Logistic Regression (1:1 nested case-control study) was used to assess if hypertension is an independent risk factor for AD, and Trend test was used to assess the relationship between blood pressure and AD. Here we demonstrate that there was a significant difference in AD prevalence among different populations stratified by systolic or diastolic blood pressure (P < 0.01). The prevalence is highest in hypertension group, and lowest in hypotension group. Multiple Logistic Regression identified high blood pressure as a risk factor for AD (OR = 1.97, 95%CI:1.09-3.54, P = 0.02). Trend test showed that there is a significant dose-response relationship between blood pressure and AD (P < 0.0002). For hypertensive AD patients, there was no significant difference in systolic blood pressure(SBP) before and after the onset of AD, but diastolic blood pressure(DBP) decreased dramatically after the onset of AD (P < 0.01); however, the result also showed that DBP decrease occurred in the non-demented group. Based on this, we think the DBP decrease is not related to AD. We further investigated whether BP values differed crossed-sectionally between the AD-patients and non-demented individuals. We found that regardless of SBP or DBP, the BP values of the AD group were all significantly higher than that of non-demented. In summary, these data suggest there is a strong relationship between hypertension and AD; however, the mechanism remains to be studied.