State of US Hospice 2026

None of this tells a family which hospice to choose. It tells them, and the people who report on this industry, what the public record actually shows about the market doing the choosing around them.

Chapter 1: The map

Where most directories count hospices by office address, this report counts them by service area: a hospice "serves" a county when at least one ZIP code in its Medicare service-area file maps to that county. This is the geography a family experiences when they search by their own address.

By that measure, 6,852 providers cover 34,827 distinct ZIP codes across 3,198 of the nation's 3,229 county-equivalents. The average service area spans 51 ZIP codes; the median is 39. The widest claims are far larger: the single largest service area in the country belongs to VITAS Healthcare of Florida at 957 ZIP codes, followed by Gentiva in Tennessee at 597. Eight of the ten widest service areas belong to for-profit chains or multi-state operators.

The density extremes define the American hospice market. At one end, Southern California: Los Angeles County is served by 1,650 hospices, San Bernardino by 999, Orange by 923, Riverside by 881. Six of the ten most-served counties in America are in Southern California; the other four lead with Harris County, Texas, at 409. At the other end, 51 counties nationwide have exactly one serving hospice, meaning one closure, one license action, or one ownership change stands between those communities and nothing. And 31 counties have no serving hospice at all.

Click to enlarge

Honesty about the deserts: they are real but remote. Twenty of the 31 zero-provider counties are Alaska boroughs and census areas; three are island municipalities in the Northern Marianas and the US Virgin Islands; the remainder are frontier counties such as Kalawao County, Hawaii (population roughly 80) and Terrell County, Texas. The sharper access story in the lower 48 is not absence but fragility: the 51 one-provider counties, and the vast middle of the distribution, where 527 counties are served by two to five providers and a single market exit meaningfully narrows a family's choices.

Chapter 2: The quality-signal gap

Medicare's primary public quality measure for hospice is the CAHPS family survey, summarized as a star rating. It requires a minimum volume of completed surveys from bereaved families. Most hospices do not have it.

Of 6,852 certified providers, 2,075 (30.3%) carry a star rating. The remaining 4,777 (69.7%) display, in CMS's phrasing, not enough survey data. A family comparing hospices sees a blank where the quality signal should be more than twice as often as they see stars.

The gap is not evenly distributed. By state, coverage ranges from 91.3% in Kentucky and 88.5% in Maryland down to 14.6% in Texas, 9.8% in Nevada, and 8.1% in California. The pattern is not random: the lowest-coverage large states are precisely the states with the largest waves of newly certified providers, a connection Chapter 3 makes explicit.

By ownership, 73.7% of nonprofit hospices carry a rating against 25.7% of for-profits. This gap should be read structurally rather than as a quality verdict: a star rating requires survey volume, survey volume requires patient census and operating history, and the nonprofit sector skews older and larger while thousands of for-profit entrants are new and small. The rating gap measures visibility, not performance.

The clearest lens is certification age. Among hospices certified before 2000, 73.1% are rated. The share falls with every subsequent cohort: 49.6% for 2000-2009, 35.1% for 2010-2014, 19.8% for 2015-2019, and 4.9% for hospices certified in 2020 or later. Where ratings do exist, average stars also drift downward by cohort, from 3.80 (pre-2000) to 3.31 (2020+), though the newest cohort's rated sample is only 133 providers and should be read accordingly.

The practical meaning for families: in exactly the markets with the most new hospices to choose among, the public quality system has the least to say about them.

Chapter 3: The new cohort

For twenty years, the United States certified roughly 200 new hospices annually. Then: 379 in 2020, 738 in 2021, 771 in 2022, before falling back to 394 in 2023 and 262 in 2024. (2025 figures are partial; the May 2026 CMS file's certification dates run only through October 2025, so the most recent counts will rise as the source catches up.)

Year certified	New hospices
2015	213
2016	212
2017	235
2018	199
2019	244
2020	379
2021	738
2022	771
2023	394
2024	262
2025 (partial)	149

The cumulative effect is a demographic reset of the industry: 2,693 of 6,852 hospices, 39.3% of the entire US market, were certified in 2020 or later.

The wave was not national. It was concentrated with unusual force in four states. California added 1,293 hospices in the period, meaning 62.7% of all California hospices are 2020-or-later entrants, and California alone accounts for 48.0% of every new hospice in America. In Nevada the new cohort is 75.0% of the state's providers; in Arizona, 50.8%; in Texas, 43.3%. No other state's new-cohort share exceeds one-third.

The new cohort also looks different from the industry it joined. Among pre-2020 hospices, nonprofits are 18.8% of the market. Among 2020-and-later entrants, they are 0.85%: 23 nonprofit hospices out of 2,693. For-profit share among the new cohort is at least 68.1%, a floor rather than an estimate, because CMS has not yet coded ownership for 765 of the newest entrants (every null-ownership provider in the country falls in the 2020+ cohort). Whatever the uncoded providers turn out to be, the near-disappearance of nonprofit entry is already visible.

This report does not attribute causes. It notes, as descriptive context, that the same four states at the center of the certification wave (California, Texas, Arizona, Nevada) were subsequently placed under CMS's enhanced-oversight provisional period for new hospices, and that California enacted a licensing moratorium in 2022. The certification curve, the geographic concentration, and the quality-signal gap of Chapter 2 are the public record's own outline of that story.

Chapter 4: California

California is one-third of this report's story on its own: 2,062 certified hospices, more than any other state, serving all 58 counties.

It is the most for-profit large market in the country: 82.9% of California hospices are for-profit against 69.2% nationally, with only 51 nonprofits statewide (2.5%). It is the epicenter of the new cohort: 62.7% of its hospices are 2020-or-later entrants, and the state accounts for nearly half of all new US hospices since 2020. And it is where the quality-signal gap is deepest: 8.1% of California hospices carry a CAHPS star rating, against 30.3% nationally, the second-lowest coverage of any state or territory with more than a handful of providers.

Its geography is the national pattern in miniature. Los Angeles County's 1,650 serving hospices lead the nation; San Bernardino, Orange, Riverside and Ventura fill four of the next five national slots. At the same time Mono County is served by a single hospice, and Humboldt, Del Norte, and a band of northern counties by fewer than five.

One verification layer exists here that exists almost nowhere else: California's public licensing file. 87.2% of the state's Medicare-certified hospices match to an active CDPH license in the June 2026 extract. The state removes non-active facilities from its consumer lookup entirely, which means absence from the file, rather than any flag on it, is the only public signal a license has lapsed. In the state at the center of the industry's newest chapter, the freshness of the public record is most of the verification a family gets.

Methodology and citation

Provider universe, service areas, quality measures and certification dates: CMS Provider Data Catalog, May 2026 refresh (Hospice Care Compare and CAHPS Hospice Survey public files). County universe: US Census Bureau 2020 FIPS county-equivalents (3,229 in the 55 covered jurisdictions). Service-area geography: Census 2020 ZCTA-to-county crosswalk; a provider serves a county when any ZIP in its CMS service-area file maps to it. California licensing: CDPH ELMS public extract, June 15, 2026.

Known limits, stated rather than smoothed: 2024-2025 certification counts are incomplete pending source updates; roughly 10% of service-area ZIPs are PO-box or territory ZIPs that map to no county; CMS has not yet coded ownership for 765 of the newest providers.

Every headline number in this report is listed, with its source, on the companion US hospice statistics page.

Cite as: HospiceAtlas, State of US Hospice 2026. hospiceatlas.com/reports. Data, queries, and county-level CSVs available on request: hello@hospiceatlas.com.

HospiceAtlas is an independent directory of all US Medicare-certified hospices, built on public Medicare (CMS) data. Not affiliated with, endorsed by, or authorized by CMS or any government agency.