How to Read HUD's Point-in-Time Homelessness Count (AHAR Part 1) — Sheltered, Unsheltered, Beds and Area Types
HUD AHAR 2025 Part 1 (principally one night in Jan 2025; six CoCs counted in Feb; excl. doubled-up, permanent housing): 745,652 homeless persons (Exhibit 1.1), -3.3% from 771,480.
Key figures (HUD AHAR 2025, Part 1: point-in-time estimates)
As of 9 November 2026, 00:55 KST (retrieved by this page at that time); latest listed on the agency page at retrieval: The 2025 Annual Homelessness Assessment Report (AHAR) to Congress, Part 1: Point-in-Time Estimates of Homelessness, from the U.S. Department of Housing and Urban Development (HUD). Report page: Posted Date: May 2026; PDF cover: May 2026. The counts were taken on one night in January 2025. Primary source: https://www.huduser.gov/portal/sites/default/files/pdf/2025-AHAR-Part-1.pdf[1][2][3]
| Indicator (one night in late January; people in ES, TH, SH or unsheltered) | Previous: January 2024 (2024 AHAR Part 1, observed) | Current: January 2025 | Change vs previous (as printed in the 2025 report) | Change vs year ago: annual count, so not a separate comparison; this column gives the 2025 table | Evidence type |
|---|---|---|---|---|---|
| All homeless persons (people) | 771,480 | 745,652 | -25,828 (-3.3%) | Exhibits 1.1 and 1.2 | HUD AHAR 2025 Part 1 (PIT count) |
| Sheltered: ES, TH and SH (people) | 497,256 | 479,332 | -17,924 (-3.6%) | Exhibits 1.13 and B1.2 | HUD AHAR 2025 Part 1 (PIT count) |
| Unsheltered (people) | 274,224 | 266,320 | -7,904 (-2.9%) | Exhibit 1.2 | HUD AHAR 2025 Part 1 (PIT count) |
| Individuals, no child in household (people) | 512,007 | 515,286 | +3,279 (+0.6%) | Exhibit B2.1 | HUD AHAR 2025 Part 1 (PIT count) |
| People in families with children (people) | 259,473 | 230,366 | -29,107 (-11.2%) | Exhibit B3.1 | HUD AHAR 2025 Part 1 (PIT count) |
| Unaccompanied youth under 25 (people) | 38,170 | 35,159 | -3,011 (-7.9%) | Exhibit B4.1 | HUD AHAR 2025 Part 1 (PIT count) |
| Veterans (people) | 32,882 | 32,495 | -387 (-1.2%) | Exhibit B5.1 | HUD AHAR 2025 Part 1 (PIT count) |
| Individuals with chronic patterns of homelessness (people) | 152,585 | 155,750 | +3,165 (+2.1%) | Exhibit B6.1 | HUD AHAR 2025 Part 1 (PIT count) |
| Year-round beds, all program types (beds) | 1,190,565 (2024 Exhibit 7-2) | 1,202,224 | not printed; +11,659 (about +1.0%, this page's calculation) | Exhibit 2.5 | HUD AHAR 2025 Part 1 (PIT count) |
Comparison basis: counts are numbers of people on one night, not rounded in the exhibits; the report's text rounds them (for example 3 percent for -3.3%). The 2024 values are observed values printed in the 2024 AHAR Part 1 (December 2024), reached from HUD's AHAR page; for the eight person-count rows, each reproduces exactly as the 2025 value minus the 2025 report's printed change (this page's calculation); the bed row has no printed change, so its change is this page's calculation. Coverage: 386 Continuums of Care (CoCs) in 2025 and 385 in 2024; American Samoa and the Northern Mariana Islands did not take part in either year.[1][4]
In three lines
1. HUD's point-in-time count found 745,652 people homeless, principally on one night in January 2025 with six CoCs counting in February (Exhibit 1.1), 3.3% fewer than 771,480 in 2024.
2. The count covers shelters, transitional housing, safe havens and unsheltered places; people doubled up or living in permanent housing programs are excluded.
3. Individuals (515,286) and chronic homelessness (155,750) rose; families with children fell 11.2% (Exhibits B2.1, B6.1, B3.1).[1]
What the count covers
Who is counted. The point-in-time (PIT) count covers people in emergency shelters (ES), transitional housing (TH), safe havens (SH) and unsheltered locations such as streets, cars and encampments. It excludes people staying with family or friends and people living in rapid rehousing (RRH), permanent supportive housing (PSH) or other permanent housing (OPH); those programs appear only as beds in the housing inventory count (HIC).[1]
Unsheltered counts are minimums. "Because of the difficulty of locating people in some of these situations and differences in local capacity to conduct the unsheltered count, the actual number of unsheltered homeless persons could be larger than reported." Communities choose among approved methods, some by full census and some by sampling, so there is no single national method.[1]
When the count happens. CoCs count during the last 10 days of January. "In 2025, HUD issued waivers to six CoCs to be able to conduct their PIT counts in February instead of the last 10 days of January due to local circumstances that made it difficult to conduct the count in January." "The CoCs that did not participate in the 2025 PIT count were American Samoa and the Northern Mariana Islands." The Augusta-Richmond County CoC in Georgia did not take part in 2025, and its 2024 data were carried over.[1]
Carried-forward data. "For the AHAR reporting, if a CoC does not conduct an unsheltered count for the reporting year, their prior year’s unsheltered data is carried forward to avoid misleading changes in the data." In 2025, 29 CoCs ran a sheltered-only count with 2024 unsheltered data carried forward; the 2024 report says: "In 2024, 22 CoCs conducted a sheltered-only count and the 2023 unsheltered count data was carried forward for these CoCs."[1][4]
Race and ethnicity. HUD changed the race and ethnicity categories in 2024, letting people choose more than one and listing Hispanic/Latina(o) as a category. "Given these changes, comparisons to prior years for race/ethnicity are not included in the report." The 2025 report does show changes between 2024 and 2025, when both years used the new categories (Exhibit B1.2).[1]
Reading in pairs
Changes by area type may reflect different CoC groupings. The 2025 report sorts 48 major-city, 64 other urban, 160 suburban and 114 rural CoCs; the 2024 report used 48, 61, 165 and 111, labelled other largely urban, largely suburban and largely rural. Subtracting the 2025 changes in Exhibit 1.13 from the 2025 counts in Exhibit 1.10 gives 2024 bases of 418,339 (major cities), 51,172 (other urban), 175,208 (suburban) and 126,761 (rural); the 2024 report printed 418,339, 45,579, 181,274 and 126,288 (Exhibit 1-13). Both sets add to 771,480 (this page's calculations). This may reflect different CoC groupings in the two reports; the report does not say so, so this remains this page's reading (see claims and unknowns).[1][4]
Shares of individuals. The report says: "About 44 percent of individuals stayed in emergency shelter, 14 percent in transitional housing, and 49 percent in unsheltered locations." Those shares add to 107%. Exhibit 3.9 gives 265,077 sheltered and 250,209 unsheltered individuals, or 51.4% and 48.6% of 515,286 (this page's calculation). Which split the sentence intends is not established.[1]
Two figures for unsheltered Hispanic people. Exhibit 1.5 prints 6,343 (24.9%) unsheltered Hispanic/Latina(o) people of any race, both in the PDF's text and on the rendered page. 6,343 is inconsistent with its printed 24.9% of 266,320 unsheltered people (6,343 would be 2.4%) and with Exhibit B1.1's 66,343, which equals 204,943 minus 138,600 (this page's calculations); why is not established.[1]
Bed growth since 2007. The executive summary says: "The total number of beds increased by 98 percent between 2007 and 2025." Section 2 says: "The total national inventory for currently homeless persons (i.e., the emergency shelter, transitional housing, and safe havens inventory) or formerly homeless persons (PSH, RRH, OPH) has increased by about 66,300 beds or 97 percent since 2007." The two statements (98 percent in the summary; about 66,300 beds or 97 percent in Section 2) cannot be reconciled from the report, because the base scope of each is not established.[1]
PSH beds for chronic homelessness. The 2025 report gives 173,080 such beds (Exhibit 2.5) and "12,533 more beds" than in 2024. The 2024 report gave 160,475 for 2024 (Exhibit 7-6), and 173,080 minus 160,475 is 12,605 (this page's calculation). Whether the 2024 figure was revised is not established.[1][4]
Rates per 10,000 people. "Nationally, about 22 of every 10,000 people were homeless on a single night." The 2024 report said about 23, and 81 per 10,000 in New York against 73 in 2025. Neither report states which population estimates it divides by, so how much of the change comes from the denominator is not established.[1][4]
Dates and next release
Dates kept apart. Count night: late January 2025. 2025 report: Posted Date: May 2026 on the report page; cover date May 2026. 2024 report: cover date December 2024. HUD also releases AHAR Part 2, with annual estimates; the AHAR page lists Part 2 reports up to the 2022 edition.[1][2][4]
Next release. The report page and the AHAR page give no date for the next Part 1 report, so it is not established.[2][3]
Verified facts
Each fact is labelled with its evidence typeEach fact in this section carries its evidence type — either cross-checked against independent sources, or confirmed from a single authoritative record.
HUD's 2025 point-in-time count found 745,652 homeless persons in the United States on a single night in January 2025, a decrease of 25,828 (3.3%) from 2024.[1] 1 source · authoritative record
Of the 745,652 people counted in January 2025, 479,332 were sheltered (emergency shelter, transitional housing or safe havens) and 266,320 were unsheltered (Exhibit 1.10).[1] 1 source · authoritative record
HUD counted 32,495 homeless veterans and 155,750 individuals with chronic patterns of homelessness in January 2025.[1] 1 source · authoritative record
HUD's 2024 report counted 771,480 people experiencing homelessness on a single night in January 2024.[4] 1 source · authoritative record
Statements and readings
Not counted as facts — each item shows what kind of statement it isFrom here on: statements not counted as facts — single-source reports, the issuing body’s own statements, and this page’s own readings, each labeled.
The 2025 report's 2024-2025 changes for other urban, suburban and rural CoCs imply 2024 bases of 51,172, 175,208 and 126,761 people, unlike the 45,579, 181,274 and 126,288 printed in 2024, consistent with CoCs being regrouped into the 2025 categories (64, 160 and 114 CoCs against 61, 165 and 111).[1][4] Stated by the issuing body itself (primary source) · not independently verified · inference from comparing Exhibits 1.10 and 1.13 of the 2025 report with Exhibit 1-13 of the 2024 report
The sentence giving 44, 14 and 49 percent for individuals adds to 107%, while Exhibit 3.9 implies 51.4% sheltered and 48.6% unsheltered.[1] Stated by the issuing body itself (primary source) · not independently verified · inference from this page's calculation with Exhibit 3.9
The report gives bed growth since 2007 as 98 percent in the executive summary and as about 66,300 beds or 97 percent in Section 2; the base scope of each figure is not established, so the two cannot be reconciled from the report.[1] Stated by the issuing body itself (primary source) · not independently verified · inference from comparing the executive summary with Section 2
Timeline
What this page could not establish
These are questions this page tried to answer and could not. The gaps are left open rather than filled with a guess.- Do the 2025 report's 2024-2025 changes by area type reflect different CoC groupings?
- What base does each bed-growth figure since 2007 (98 percent; 66,300 beds or 97 percent) use?
- Which population estimates are used for the rates per 10,000 people?
- When will the next Part 1 report be posted?
- Model
- claude-opus-5
- Time
- 11/09/2026, 00:55
- Body characters
- 8,414
- Sources
- 4 sources adopted
- Model
- ChatGPT (checked sources on the web)
- Time
- 11/09/2026, 01:25
- Verdict
- Revision required
Show revision history (3)
| 11/09/2026, 09:00 | First authored (claude-opus-5) | Created |
| 11/09/2026 | First version. Read in full the 2025 AHAR Part 1 (125 pages), its report page, the AHAR page and the 2024 AHAR Part 1 (117 pages) for observed 2024 values; quotations are complete sentences matched against the source text. | Updated |
| 11/09/2026 | Revised after review: the bed-growth paragraph and claim now say only that the summary's 98 percent and Section 2's 66,300 beds or 97 percent cannot be reconciled because their base scope is not established; Exhibit 1.5's 6,343 was confirmed as printed and is described as inconsistent with its 24.9% and with Exhibit B1.1's 66,343; the area-type paragraph says the changes may reflect different CoC groupings; the exact-reproduction statement is limited to the person-count rows; the summary says the count was taken principally on one night in January 2025, with six CoCs counting in February; source notes give page counts read in full from the retrieved copies. | Updated |
Frequently asked
How many people were homeless in the United States in January 2025?
745,652 people on a single night in January 2025 (Exhibit 1.1), 25,828 or 3.3% fewer than the 771,480 counted in January 2024. The count covers shelters, transitional housing, safe havens and unsheltered places.[1]
Does the count include people staying with friends or family?
No: the 745,652 count excludes people doubled up or couch surfing. HUD's definition of homelessness used for the count covers only shelters, transitional housing, safe havens and places not meant for habitation.[1]
Are people in permanent supportive housing counted as homeless?
No: the 408,167 permanent supportive housing beds in 2025 (Exhibit 2.5) are counted as inventory, and the people living in them are not in the point-in-time count. The same applies to rapid rehousing and other permanent housing.[1]
How many homeless veterans were counted?
32,495 in January 2025, 387 (1.2%) fewer than in 2024 and 55.7% fewer than in 2009 (Exhibit B5.1). About four in ten were unsheltered.[1]
Which state has the highest rate of homelessness?
New York, with 73 of every 10,000 people counted as homeless in 2025 (Exhibit 1.7). Appendix A also shows 73 in every 10,000 for the District of Columbia, which is not a state.[1]
How are the unsheltered numbers estimated when a CoC skips its street count?
In 2025, 29 CoCs counted only sheltered people, and their 2024 unsheltered counts were carried forward. HUD says this avoids misleading changes in the data.[1]
Official links
- Official HUD User — AHAR data and reports
Sources
- [1] The 2025 Annual Homelessness Assessment Report (AHAR) to Congress, Part 1: Point-in-Time Estimates of Homelessness, PDF primary
- [2] 2025 AHAR: Part 1 - PIT Estimates of Homelessness in the U.S. (HUD User report page) primary
- [3] Annual Homelessness Assessment Report (AHAR) Data and Reports (HUD User) primary
- [4] The 2024 Annual Homelessness Assessment Report (AHAR) to Congress, Part 1: Point-in-Time Estimates of Homelessness, PDF primary