Updated August 6, 2026. Quick answer: 63 million Americans are family caregivers — about one adult in four, and a 45% increase in a decade. The money statistics are the ones that get quoted least and matter most: over a third have stopped saving, and 23% are in debt because of it.
The numbers
63 million Americans are family caregivers. The 2025 edition of the longest-running survey in this field counts 63 million family caregivers. Roughly 1 in 4 American adults. Of those, 59 million care for adults and 4 million for children under 18 with an illness or disability. [Caregiving in the US 2025]
That is a 45% increase in a decade. A “45 percent increase, or nearly 20 million more caregivers, over the past decade”. This is the number that reframes the topic: caregiving is not a stable background condition, it is the fastest-growing unpaid job in the country. [Caregiving in the US 2025]
44% provide high-intensity care. “Forty-four percent of caregivers report providing high-intensity care” — not occasional help, but the level that displaces paid work. [Caregiving in the US 2025]
30% have been doing it for five years or more. “30 percent having been in the role for five years or more.” Long enough for a career gap to become permanent and for retirement contributions missed in those years to be unrecoverable. [Caregiving in the US 2025]
Over a third have stopped saving. “A little over one-third report that they have stopped saving”. This is the statistic with the longest tail: contributions not made in your fifties are the ones with the least time left to compound. [Caregiving in the US 2025]
24% have exhausted short-term savings; 13% have tapped long-term savings. “24 percent have exhausted personal short-term savings and 13 percent have tapped into long-term savings.” Tapping long-term savings is the one that usually carries a tax bill and sometimes a penalty on top of the loss. [Caregiving in the US 2025]
23% are in debt because of it. “Twenty-three percent also report being in debt.” [Caregiving in the US 2025]
1 in 5 report fair or poor health themselves. “nearly 1 in 5 (20 percent) report fair or poor health.” The caregiver is frequently the next person to need care, which is why planning that treats them as an unlimited resource fails twice. [Caregiving in the US 2025]
What the numbers mean for money
Put the duration and the savings figures together and the cost stops looking like an expense and starts looking like a hole in a retirement plan. Five or more years of caregiving — the position 30% are in — overlaps precisely with the highest-earning, highest-contribution years of a career. Contributions not made in your fifties have the least time left to compound, which is why “stopped saving” is a more expensive sentence than “spent savings”.
Two structural responses exist and both are underused. A family caregiver agreement turns informal care into a documented arrangement — which matters for fairness between siblings and for Medicaid look-back, where undocumented transfers to a caring child are treated as gifts. And the caregiver child exemption can protect the house itself, but only where the caregiving was documented and met the rules before it mattered.
The 20% reporting fair or poor health is the figure that should change planning behaviour. A plan that treats the family caregiver as an unlimited free resource has a single point of failure who is, statistically, in worse health than average. What the paid alternative actually costs is at long-term care statistics and long-term care planning.
How we read this, and what is missing
One source, named, with nothing averaged across studies. Every figure above is from Caregiving in the US 2025, published by AARP and the National Alliance for Caregiving, read from AARP’s own page on August 6, 2026. Caregiving statistics vary wildly between surveys because they define “caregiver” differently; blending them produces numbers that belong to no study at all.
The hours-per-week figure is not here, and that is a gap. The time-use data that would support it comes from the Bureau of Labor Statistics, whose site does not serve our requests. We would rather name the hole than fill it from a secondary source quoting a number we have not read.
More of our data work is at statistics. See methodology and corrections — this report is republished periodically, and when the next edition lands these figures get updated with the new date shown.
All the numbers, kept current. This page uses 4 figures from our claims register — every figure we track is on one page, each with the year it applies to and a plain statement of what makes it move.
What a caregiver can actually claim depends on where they work: paid family leave by state, all 51 jurisdictions, with the caregiver-leave detail national summaries flatten.