US Cities by Homelessness Rate, 2026
Major U.S. urban service areas ranked by homelessness rate
This ranking compares selected major U.S. city-centered homelessness service areas by the number of people experiencing homelessness per 10,000 residents. The numerator is the total Point-in-Time homelessness count reported through HUD’s Continuum of Care system. The denominator is the closest matching resident population for the same city, county, multi-county or metropolitan geography.
The 2026 label refers to the publication snapshot compiled in May 2026. The underlying homelessness counts are HUD’s January 2024 PIT data, the latest fully published AHAR Part 1 dataset available on HUD User for consistent national comparison at the time of compilation. Preliminary 2025 local estimates are not mixed into the table because they did not have the same federal AHAR Part 1 quality-review status.
This is not a strict city-proper ranking. Some Continuums of Care align closely with a municipality, while others cover a county or regional homelessness-service area. The table keeps those geography labels visible so that New York City, Boston, Los Angeles City & County, Seattle / King County and Houston regional figures are not read as identical municipal units.
New York City CoC ranks first among the listed urban service areas, measured per 10,000 residents.
New York City also has the largest total PIT count in the table.
The list covers 48 selected large city, county and metro CoCs where the PIT count can be paired with a compatible population denominator.
Rates are rounded to one decimal place. The national 2024 PIT rate was roughly 23 people per 10,000 U.S. residents.
What the homelessness rate measures
The rate converts a one-night homelessness count into a population-adjusted measure. A place with a high absolute count does not automatically rank highest by rate; the resident population denominator changes the result. This makes the metric useful for comparing the relative pressure on shelter systems, outreach capacity, permanent housing pipelines and local service coordination.
The measure is not a complete count of every person lacking stable housing. PIT counts focus on people in emergency shelter, safe haven, transitional housing and unsheltered locations on one night in January. People doubled up with relatives, staying temporarily with friends or moving between informal arrangements are often outside the PIT definition.
Top 10 urban service areas by homelessness rate
The upper tier is not one uniform pattern. New York City and Boston have large sheltered populations, while San Francisco, Portland, Seattle, Long Beach and Los Angeles combine severe housing-cost pressure with substantial unsheltered homelessness. Chicago’s 2024 count rose sharply, placing it high on a population-adjusted basis.
| Rank | Urban service area | Rate per 10,000 | Total PIT count |
|---|---|---|---|
| 1 | New York City CoC | 165.3 | 140,134 |
| 2 | San Francisco CoC | 100.6 | 8,323 |
| 3 | Portland / Multnomah County CoC | 92.2 | 7,384 |
| 4 | Boston CoC | 90.1 | 5,898 |
| 5 | District of Columbia CoC | 80.0 | 5,616 |
| 6 | Long Beach CoC | 75.1 | 3,376 |
| 7 | Los Angeles City & County CoC | 73.9 | 71,201 |
| 8 | Seattle / King County CoC | 72.5 | 16,868 |
| 9 | Chicago CoC | 69.2 | 18,836 |
| 10 | Oakland, Berkeley / Alameda County CoC | 58.0 | 9,450 |
Rates use matched geography denominators. A city CoC, county CoC and metro CoC should not be interpreted as identical municipal units.
Chart: highest rates among the selected urban areas
The first fifteen entries show how sharply the rate falls after New York City. The comparison also shows why rates and total counts should be read together: Los Angeles and Seattle have much larger counts than some smaller areas, but their rates depend on broader city-county or countywide denominators.
Methodology
The ranking uses a rate per 10,000 residents to compare homelessness across large U.S. urban service geographies. Total PIT counts come from HUD’s federal Continuum of Care reporting system for January 2024. Population denominators are matched to the geography of each CoC as closely as possible, using city, county, multi-county or metropolitan population estimates from the U.S. Census Bureau.
Indicator formula
Homelessness rate = total PIT homelessness count ÷ matched resident population × 10,000.
Data period
The homelessness count is the January 2024 PIT snapshot, published in HUD’s 2024 AHAR Part 1 release. The page is labeled as a 2026 publication snapshot compiled in May 2026.
Geography rule
City CoCs use city population where appropriate. County and regional CoCs use county, multi-county or metro denominators to avoid mixing a broad PIT count with a narrow city population.
Rounding
Rates are rounded to one decimal place. PIT counts are shown as whole persons. Population denominators are used for calculation and summarized separately for key rows.
The selection includes large urban and city-centered CoCs where the PIT count can be paired with a compatible denominator without treating a countywide count as if it were a city-only count. That is why the table has 48 rows rather than being labeled as a Top 50 or Top 100 list.
PIT data should be interpreted as a one-night estimate. Counts may miss hidden homelessness, people avoiding contact with outreach teams and residents doubled up outside the HUD homelessness definition. Unsheltered counts are especially sensitive to field coverage, weather, encampment visibility and local methodology. HUD notes that CoC-level information receives limited federal data-quality review but is not independently verified in every detail by HUD.
Full ranking: 48 selected urban service areas
The table ranks selected major U.S. city-centered CoCs and large urban service geographies by homelessness rate per 10,000 residents. The original rank remains fixed, while the controls change the visible subset and order.
| Rank | Urban service area | Rate per 10,000 | Total PIT count |
|---|---|---|---|
| 1 | New York City CoC | 165.3 | 140,134 |
| 2 | San Francisco CoC | 100.6 | 8,323 |
| 3 | Portland / Multnomah County CoC | 92.2 | 7,384 |
| 4 | Boston CoC | 90.1 | 5,898 |
| 5 | District of Columbia CoC | 80.0 | 5,616 |
| 6 | Long Beach CoC | 75.1 | 3,376 |
| 7 | Los Angeles City & County CoC | 73.9 | 71,201 |
| 8 | Seattle / King County CoC | 72.5 | 16,868 |
| 9 | Chicago CoC | 69.2 | 18,836 |
| 10 | Oakland, Berkeley / Alameda County CoC | 58.0 | 9,450 |
| 11 | Atlanta CoC | 56.1 | 2,867 |
| 12 | San Jose / Santa Clara City & County CoC | 54.0 | 10,394 |
| 13 | Albuquerque CoC | 48.9 | 2,740 |
| 14 | Metropolitan Denver CoC | 47.7 | 14,281 |
| 15 | Honolulu City and County CoC | 45.3 | 4,497 |
| 16 | Fresno City & County / Madera County CoC | 36.8 | 4,305 |
| 17 | Philadelphia CoC | 33.5 | 5,191 |
| 18 | Las Vegas / Clark County CoC | 32.9 | 7,906 |
| 19 | San Diego City and County CoC | 32.3 | 10,605 |
| 20 | Minneapolis / Hennepin County CoC | 29.7 | 3,866 |
| 21 | Nashville / Davidson County CoC | 29.3 | 2,094 |
| 22 | Bakersfield / Kern County CoC | 28.9 | 2,669 |
| 23 | Baltimore CoC | 28.3 | 1,600 |
| 24 | Oklahoma City CoC | 26.2 | 1,838 |
| 25 | Louisville-Jefferson County CoC | 22.1 | 1,728 |
| 26 | Austin / Travis County CoC | 21.8 | 2,975 |
| 27 | Tulsa City & County CoC | 20.3 | 1,389 |
| 28 | Phoenix, Mesa / Maricopa County CoC | 20.2 | 9,435 |
| 29 | Tucson / Pima County CoC | 20.1 | 2,142 |
| 30 | New Orleans / Jefferson Parish CoC | 19.5 | 1,454 |
| 31 | Kansas City regional CoC | 18.8 | 2,181 |
| 32 | Charlotte / Mecklenburg CoC | 17.8 | 2,095 |
| 33 | Columbus / Franklin County CoC | 17.6 | 2,380 |
| 34 | Indianapolis CoC | 17.5 | 1,701 |
| 35 | Omaha, Council Bluffs CoC | 16.4 | 1,609 |
| 36 | San Antonio / Bexar County CoC | 16.2 | 3,398 |
| 37 | Colorado Springs / El Paso County CoC | 14.9 | 1,146 |
| 38 | Miami-Dade County CoC | 14.1 | 3,800 |
| 39 | Dallas City & County / Irving CoC | 14.1 | 3,718 |
| 40 | Tampa / Hillsborough County CoC | 12.0 | 1,893 |
| 41 | Fort Worth / Arlington / Tarrant County CoC | 11.3 | 2,463 |
| 42 | Jacksonville-Duval, Clay Counties CoC | 10.5 | 1,339 |
| 43 | El Paso City & County CoC | 10.5 | 913 |
| 44 | Milwaukee City & County CoC | 9.7 | 885 |
| 45 | Memphis / Shelby County CoC | 8.6 | 784 |
| 46 | Raleigh / Wake County CoC | 8.1 | 992 |
| 47 | Virginia Beach CoC | 6.9 | 311 |
| 48 | Houston, Pasadena, Conroe regional CoC | 5.8 | 3,280 |
Data basis: HUD 2024 AHAR Part 1 and HUD Exchange CoC population reports for PIT counts; U.S. Census Bureau city, county and metropolitan population estimates for denominators. Rates are analytical estimates, rounded to one decimal place.
Population denominators used for key rows
The rate is only meaningful when the PIT count and resident population cover comparable geography. The examples below show the calculation basis for the highest-impact rows. Population denominators are rounded to make the calculation transparent; the table rate remains the authoritative rounded result.
| Urban service area | PIT count | Approx. denominator | Rate per 10,000 |
|---|---|---|---|
| New York City CoC | 140,134 | 8.48 million city residents | 165.3 |
| San Francisco CoC | 8,323 | 827,000 city residents | 100.6 |
| Portland / Multnomah County CoC | 7,384 | 801,000 county residents | 92.2 |
| Boston CoC | 5,898 | 655,000 city residents | 90.1 |
| District of Columbia CoC | 5,616 | 702,000 district residents | 80.0 |
| Los Angeles City & County CoC | 71,201 | 9.63 million countywide residents | 73.9 |
| Seattle / King County CoC | 16,868 | 2.33 million county residents | 72.5 |
| Chicago CoC | 18,836 | 2.72 million city residents | 69.2 |
| Oakland, Berkeley / Alameda County CoC | 9,450 | 1.63 million county residents | 58.0 |
| Houston, Pasadena, Conroe regional CoC | 3,280 | 5.66 million regional residents | 5.8 |
The denominator examples explain why the ranking should be read as an urban service-area comparison. A countywide or regional CoC is not converted into a city-only rate.
Insights from the distribution
Top tier
The highest rates combine two forces: large PIT counts and tight denominators. New York City’s exceptional rate reflects a very large shelter-system count relative to the city population. San Francisco, Portland, Long Beach, Los Angeles, Seattle and Alameda County show how high housing costs and visible unsheltered homelessness can push West Coast service areas toward the top even when their absolute counts differ widely.
Middle range
The middle of the table includes county and metro CoCs where homelessness is substantial but spread across broader resident populations. San Diego, Las Vegas, Minneapolis, Nashville and Austin illustrate why a high count is not enough to rank at the top when the denominator includes a larger county or regional base.
Lower listed rates
The lower rows are not necessarily places without housing stress. In Houston, Tarrant County, Raleigh and Virginia Beach, the rate is lower because the listed CoC count is measured against a large denominator or because the reported PIT count is comparatively small. Local service access, hidden homelessness and counting methods still matter.
Regional pattern
The West is heavily represented in the top half, especially California and the Pacific Northwest. The Northeast appears near the top through New York City, Boston, Washington, Philadelphia and Baltimore. The South has many listed service areas, but most are concentrated in the middle and lower portions of this specific population-adjusted table.
What this means for readers
For residents, the rate helps separate the scale of visible homelessness from the size of the local population. A high rate signals that homelessness is not just a large-city issue in absolute terms; it is a severe per-capita service challenge. That matters for public expectations around shelter availability, outreach coverage, sanitation, emergency response, behavioral-health capacity and permanent housing placement.
For analysts and local officials, the ranking is a screening tool. It does not prove which city has the best or worst policy response, but it shows where deeper local diagnosis is needed. A high rate should be examined alongside rent burden, vacancy rates, eviction filings, shelter utilization, permanent supportive housing production, inflow into homelessness and exits to stable housing.
For businesses and civic groups, the data provides a practical sense of local pressure on downtowns, transit corridors, hospitals, libraries and nonprofit service networks. Because the PIT count is a one-night snapshot, the ranking should guide questions rather than replace local operational data.
FAQ
Is this a ranking of city governments?
No. It is a ranking of selected city-centered homelessness service geographies. Some rows are close to city boundaries, while others cover counties or regional Continuums of Care. That distinction is central to interpreting the rates.
Why use a rate per 10,000 residents?
The rate adjusts the PIT count for population size. Without this adjustment, the largest cities would dominate the table simply because they have more residents.
Why does New York City rank so high?
New York City has both the largest PIT count in the table and a city-level denominator. Its extensive shelter system also means many people experiencing homelessness are counted in sheltered settings.
Why are 2025 preliminary local counts not used?
The table uses the latest federally published AHAR Part 1 dataset available for consistent CoC-level comparison at the compilation date. Preliminary local 2025 counts may be useful for local monitoring, but they are not equivalent to the HUD AHAR Part 1 national release.
Does a lower listed rate mean homelessness is not a serious issue?
No. A lower rate can result from a broad denominator, local counting conditions, fewer unsheltered people visible during the count, or genuine differences in homelessness prevalence. The rate should be read together with local shelter demand, housing costs and service capacity.
Are doubled-up households included?
Usually not. The HUD PIT definition focuses on people in sheltered homelessness programs and unsheltered locations on one night. Many people in unstable informal housing arrangements are outside the PIT count.
Sources
-
HUD User — 2024 AHAR Part 1: PIT Estimates of Homelessness in the U.S.
Primary federal source for national, state and CoC-level January 2024 PIT estimates used as the homelessness-count basis. -
HUD User — Annual Homelessness Assessment Report data and reports
Used to verify the AHAR publication cycle and the latest federally published Part 1 PIT report available at the compilation date. -
HUD Exchange — CoC Homeless Populations and Subpopulations Reports
Used for Continuum of Care PIT count detail and HUD’s note on limited data-quality review of CoC-submitted information. -
HUD Exchange — Point-in-Time Count and Housing Inventory Count
Used for PIT/HIC definitions, submission context and the single-night count framework. -
U.S. Census Bureau — Population Estimates Program
Main population-estimate source family used to select city, county and metropolitan denominators. -
U.S. Census Bureau — City and Town Population Totals
Used for city-level denominators where the CoC geography closely matches a municipality. -
U.S. Census Bureau — County Population Totals
Used for county-level denominators where the CoC covers a county or county-centered service area. -
U.S. Census Bureau — Metropolitan and Micropolitan Statistical Area Population Totals
Used for broader regional denominators when a CoC is best matched to a metro or multi-county service geography.
