ACL Forecasts $39.7 Million for Dementia in FY2027: The Respite Program Gets Fewer, Larger Awards, and States Get a Second Shot

On October 19, 2026, the Administration for Community Living posted ten FY2027 forecasts to Grants.gov in a single batch. Three belong to the Alzheimer's Disease Programs Initiative (ADPI), ACL's main discretionary program for building dementia-capable home and community-based services. Together they forecast $39.7 million:

All three are 36-month cooperative agreements, all three flag cost sharing as required, and all three list estimated award dates of July 30, 2027. The same batch also included two caregiver and access forecasts that dementia applicants should track (covered below).

These are forecasts, not open solicitations. The dates are ACL's estimates, and in at least one case the estimate is obviously a placeholder. Still, a forecast posted four to seven months ahead is the earliest public signal ACL gives, and the gap between this cycle and the last one is large enough to change who should apply.

The Respite Program Changed Shape

The biggest change is in dementia-specific respite.

ACL launched the State and Territory Programs for Dementia-Specific Respite as a new ADPI component in July 2026 (HHS-2026-ACL-AOA-ADPI-0036). That NOFO offered up to $2 million per year, split across up to 30 awardees, for three-year projects. Applications were due August 3, 2026, and awards were expected September 30. If every slot had been filled, the average award would have been about $67,000 a year, which is enough for a pilot in a few Area Agencies on Aging and not much more.

The FY2027 forecast lists $23 million for 15 awards, with a ceiling of $4 million and a floor of $500,000. That is half as many awards and roughly ten times the money. Read the forecast as written: this is no longer a seed grant spread across most of the states. It is a demonstration program sized for a State Unit on Aging (SUA) to fund respite across a real share of its AAA network.

The program design is unchanged. Funds go to SUAs. SUAs pass them to AAAs and other aging network providers, who deliver dementia-specific respite "comparable to those provided under Title III-E, the National Family Caregiver Support Program." Providers must have dementia-specific training. Eligible caregivers include unpaid family caregivers of people with a probable diagnosis of Alzheimer's disease or a related dementia, including caregivers of people under 60 with younger-onset dementia who otherwise meet Title III-E rules. That last group is often left out of Older Americans Act services because of the age-60 threshold, and ACL wrote it into the program on purpose.

We could not find a public ACL announcement of FY2026 respite awards as of October 28. If your state applied in August, assume the FY2027 competition is open to you unless the final NOFO says otherwise. If your state did not apply, the larger awards make it worth a second look.

What an SUA should do now:

1. Pick the AAAs. A $1 million to $4 million award only makes sense if you can name the AAAs that will deliver services, the counties they cover, and their current Title III-E respite waitlists. Get letters of commitment before the NOFO posts.
2. Decide who does the training. The forecast requires providers with "appropriate dementia-specific training." Name the curriculum and the trainer, whether that is a university partner, an Alzheimer's Association chapter, or a Dementia Friendly America affiliate.
3. Line up the cost share. Prior ADPI NOFOs, including the FY2026 respite notice, required a 25 percent non-federal share. On a $4 million request, that means finding about $1.33 million. State general-fund respite lines, Medicaid waiver administrative match, and in-kind AAA staff time are the usual sources. Price them now.
4. Write the younger-onset pathway down. Explain how a 52-year-old with early-onset Alzheimer's and their spouse get referred, screened, and served. Reviewers will look for it because the forecast calls it out.

States and Communities: A Second Shot

The ADPI States and Community Grants forecast continues ACL's longest-running dementia program. One NOFO has two options:

• Option A, Grants to States. Open to the state or territorial agency designated for dementia capability. Projects must build a dementia-capable statewide home and community-based services (HCBS) system with Single Entry Point/No Wrong Door access for people living with dementia and their family caregivers, plus a sustainable set of dementia-capable services. Only states without an active ACL ADPI state grant are eligible. States that "have not benefited from ADPI/ADI-SSS/ADSSP grants since before 2014" get priority consideration after review.
• Option B, Grants to Communities. Open to public or private community-based organizations that already operate inside a dementia-capable HCBS system and can name specific gap areas to fill.

No entity can apply under both options, and no entity can hold more than one ADPI state or community grant at a time.

The money is close to last year's. The FY2026 version (HHS-2026-ACL-AOA-ADPI-0002) was listed at about $10 million for 12 awards of $850,000 to $1 million, with a March 17, 2026 deadline. FY2027 moves to $15 million and 17 awards in the same range. That is five more slots.

The estimated due date of March 3, 2027 comes two days after the estimated posting date of March 1. That is almost certainly a placeholder. Last year's cycle gave applicants about eight weeks, and you should expect something similar. Do not read the forecast as a promise of a two-day window. Do read it as a sign that the NOFO is likely to post in late winter, with a deadline in spring.

Who should apply: For Option A, the best fit is a state that has never held an ADPI grant, or last held one before 2014. That is the stated priority, and it carries real weight when ACL makes final decisions on a tight scoring band. For Option B, the best fit is a community-based organization that can show it is part of an existing dementia-capable system. A standalone memory café is not enough. You need referral agreements with your AAA or ADRC and a clear gap you will fill: rural reach, a language community, people with intellectual and developmental disabilities who develop dementia, or people living alone with dementia.

Indian Country: Six Awards, Two Tracks

The Dementia Capability in Indian Country forecast is small: $1.7 million for six awards of $200,000 to $300,000. It is open to federally recognized tribes, tribal organizations, and consortia that already provide home and community services in tribal communities. Option A is for tribes that held a previous ADPI grant and want to expand that work. Option B is for tribes new to ADPI that will add dementia services to existing systems. Applications must show tribal leadership support and the authority to carry out the work.

New tribes should start early on the leadership-support requirement. A tribal council resolution can take a full meeting cycle or two. If your council meets monthly and the NOFO posts March 1 with a May 3 deadline, start the resolution process in January. Title VI Native American aging programs are the natural lead, because the forecast requires that the applicant is "actively providing" home and community services now.

Two Related Forecasts in the Same Batch

Two non-ADPI forecasts in the October 19 batch belong in the same planning conversation:

• Advancing State Implementation of the National Strategy to Support Family Caregivers (HHS-2027-ACL-AOA-CGNS-0039): $2.96 million, six awards of $450,000 to $490,000, two-year projects, estimated due July 22, 2027. Only SUAs that administer Title III-E can apply. Grantees get up to six months for planning and must turn in a sustainability plan at the end of it.
• No Wrong Door System Modernization (HHS-2027-ACL-CIP-NWGV-0005): $4.5 million, ten awards of exactly $450,000, no cost share, estimated due March 18, 2027. It names "emerging technologies including artificial intelligence" as a way to improve access.

For a state aging agency, these and ADPI fit together. ADPI Option A requires No Wrong Door access for people with dementia. The NWD modernization grant pays to rebuild that front door. The caregiver strategy grant pays for the cross-agency planning that dementia respite needs. A state that plans all three together can write proposals that support each other instead of three unrelated ones.

The Risk Behind the Forecasts

These forecasts come from an agency whose future is uncertain. HHS proposed in 2025 to break up ACL and spread its programs across ACF, ASPE, and CMS. It closed regional offices and cut nearly half of ACL's staff. Congress declined that reorganization. The FY2027 budget request tries again, proposing to fold ACL and ACF into a new Administration for Children, Families, and Communities at $28.7 billion, about $7 billion below FY2026. House appropriators have not adopted the merger. According to NCOA's September 2026 tracker, the House bill level-funds ADPI, which is paid through the Prevention and Public Health Fund, and level-funds Title III-E. The Senate has not released its aging numbers.

The government is running on a continuing resolution through December 11, 2026. A full-year CR at FY2026 levels would likely keep these programs alive at roughly forecast scale. A merger enacted in an omnibus could delay NOFOs while grant systems and staff move. Plan for the forecast, and expect the NOFO to arrive late.

What to Do Before the NOFOs Post

• Subscribe to all three forecast numbers on Grants.gov so you get notified when the forecast becomes a posted NOFO.
• SUAs: decide in November which of the three state-eligible opportunities (respite, ADPI Option A, NWD) you will lead, and which you will support through a partner.
• Community organizations: get a letter from your AAA or ADRC that describes the existing dementia-capable system you operate within. Option B depends on it.
• Tribes: start the council resolution process and document current home and community services.
• Everyone: set aside the 25 percent match in your FY2027 budget now, before the money goes elsewhere.

When the NOFOs post, Granted can map each review criterion to a section outline so your team starts writing from the scoring rubric.