STUDY QUESTION: Did the Medicare Alzheimer's Disease Demonstration with its case management and community service waivers affect the use of community-based long-term care services among people with dementia and their primary caregivers? DATA SOURCES: Baseline and periodic caregiver interviews. Measures include client and caregiver attributes and self-reported service use. STUDY DESIGN: The demonstration randomly assigned voluntary applicants into treatment and control groups. Treatment group cases were eligible for case management and for up to $699 per month in community care benefits. The actual monthly entitlement varied among the eight demonstration communities due to regional cost and inflation adjustments over time. Analyses are for the year after enrollment. DATA COLLECTION: Analyses are of cases surviving six months or more in the community after enrollment (n = 5,209). Cases received baseline and semi-annual assessments. PRINCIPAL FINDINGS: The intervention of case management and community service reimbursement had a strong, consistent, and positive effect on the likelihood of using home care (including homemaker/chore services, personal care services, companion services) and adult day care. Treatment group clients were at least twice as likely as control group clients to be using any of the four community-based services. A similar, but less pervasive effect was achieved with caregiver training and support group participation. Reimbursement provided by the demonstration's Medicare waiver was generally not sufficient to exceed the level of control group service acquired through private payment. CONCLUSIONS: Reimbursement levels within the demonstration may have enabled more individuals to purchase some services, but they were not sufficient to increase the average level of use over those in the control group. No consistent differences between demonstration models were found in service use likelihood or average use among users.