Annual Homeless Assessment Report: Structure, Data, and Deadlines

The Annual Homeless Assessment Report is the yearly report that the U.S. Department of Housing and Urban Development delivers to Congress estimating how many people are experiencing homelessness across the country, who they are, and how the housing and shelter system is serving them. Federal law under 42 U.S.C. § 11360a requires HUD to submit it, and the January 2024 count found roughly 771,480 people experiencing homelessness on a single night, a 10 percent increase over the prior year.¹ The data drives billions of dollars in competitive funding decisions for local communities.

What Federal Law Requires HUD to Report

The statutory authority sits in 42 U.S.C. § 11360a, part of the McKinney-Vento Homeless Assistance Act as amended by the HEARTH Act. It directs the Secretary of HUD to submit an annual report assessing the number of people experiencing homelessness across more than a dozen specific categories: people who are chronically homeless, veterans, unaccompanied youth, families with children, domestic violence survivors, people with disabilities, and people with substance use disorders or serious mental illness, among others.² The statute also requires HUD to gather this data from every Continuum of Care in the country and from any other sources the Secretary considers appropriate.

A separate statute, 42 U.S.C. § 11313, requires the U.S. Interagency Council on Homelessness to prepare its own annual report to the President and Congress on the nature and extent of homelessness and federal accomplishments in addressing it.³ The two reports serve different purposes but rely on much of the same underlying local data.

The Two Parts of the Report

HUD releases the AHAR in two separate parts, each built from a different data source and answering a different question.¹

Part 1: Point-in-Time Estimates

Part 1 draws on the Point-in-Time (PIT) count conducted each January and provides a snapshot of how many people were homeless on a single night. It breaks the total down by sheltered and unsheltered status, geography, and demographics. The 2024 Part 1 report was released in late December 2024, though release timing has varied year to year. Part 1 also includes the Housing Inventory Count, which catalogs every bed and unit available in the homeless services system.

Every Continuum of Care must conduct the PIT count during the last 10 days of January.⁴ People staying in emergency shelters and transitional housing are counted as sheltered. People sleeping in places not meant for habitation, such as cars, parks, sidewalks, or abandoned buildings, are counted as unsheltered.⁵ Federal regulations require the full count at least every two years, but HUD by Notice requires the sheltered component annually and the unsheltered component at a minimum in odd-numbered years. The counts collect demographic data including age, race, ethnicity, gender, veteran status, and whether a person has a disabling condition, feeding directly into the assessments that 42 U.S.C. § 11360a requires.²

The Housing Inventory Count runs alongside the PIT count and captures emergency shelters, transitional housing, rapid re-housing, permanent supportive housing, and other permanent housing programs. Emergency shelter capacity is broken out into year-round beds available every night, seasonal beds available during a defined higher-demand period, and overflow beds opened temporarily when demand exceeds planned capacity. A community with no fixed number of overflow beds may report the number actually occupied on the night of the count.⁶ Comparing the two datasets exposes gaps: if a community counts 2,000 people experiencing homelessness but its inventory holds 1,200 beds, that 800-bed gap becomes a data point that drives federal funding decisions.

Part 2: Longitudinal Systems Analysis

Part 2 takes a longer view. It uses Longitudinal Systems Analysis (LSA) data submitted by Continuums of Care to estimate the total number of people who experienced homelessness over a full year, not just a single night. The LSA data covers how long people stayed homeless, how they moved through shelters and housing programs, and where they went when they exited. It also isolates populations with distinct needs, such as veterans and people experiencing chronic homelessness. The LSA export consists of 12 CSV files covering individual demographics, project funding, and inventory data, giving HUD a granular picture that a one-night count cannot provide.⁷

Who Counts as Homeless Under the Federal Definition

HUD’s definition is broader than many people assume, and it shapes what the AHAR captures. It covers four categories:

People who are literally homeless: living in shelters, transitional housing, or places not meant for habitation.
People at imminent risk of homelessness: those who will lose their housing within 14 days and have no subsequent residence or resources to obtain one.
Families or unaccompanied youth who qualify as homeless under other federal statutes and have experienced housing instability.
People fleeing or attempting to flee domestic violence who have no other safe housing.⁸

The PIT count primarily captures the first category. The LSA data behind Part 2 can capture a broader range of people who accessed services throughout the year.

How Local Data Reaches HUD

The AHAR’s numbers are only as good as the local data feeding them. Two systems do that work: Continuums of Care and the Homeless Management Information System.

Continuums of Care

Continuums of Care are the local planning bodies that coordinate homeless services and submit data to HUD. Each covers a specific geographic area, from a single city to a multi-county region, and typically includes nonprofits, local government agencies, hospitals, and other stakeholders. Federal regulations at 24 CFR Part 578 govern how they operate and what they must report.⁹ If HUD determines that a CoC or one of its grant recipients has failed to carry out its responsibilities, consequences range from corrective action plans to suspension of payments to termination of the remaining grant. Timely and accurate data submission also factors into how HUD scores each community’s application in the annual CoC Program Competition, which determines federal funding levels.

HMIS Standards

The Homeless Management Information System is the software platform providers use to record client-level data in real time. Every CoC must select an HMIS that complies with HUD’s data collection, management, and reporting standards.¹⁰ The HEARTH Act made HMIS participation a statutory requirement for recipients of CoC Program and Emergency Solutions Grants funding. Standardized software following uniform rules is what lets HUD aggregate data from hundreds of communities into a single national report.

HUD’s FY 2026 HMIS Data Standards, effective October 1, 2025, require every participating project to record a set of Universal Data Elements for every client: name; Social Security number (full or partial); date of birth; race and ethnicity; sex (new as a standalone element for FY 2026); veteran status for all adults; disabling condition; and relationship to head of household.¹¹ Additional fields track each person’s movement through the system: project start and exit dates, prior living situation, exit destination, and housing move-in date.

Privacy Rules and the Survivor Carve-Out

Federal rules do not require client consent before asking for or entering data, because projects are required by their funders to collect it. Any use or disclosure not described in the CoC’s privacy notice does require consent, and clients can refuse to provide information without being denied services.¹¹ Survivors of domestic violence receive stronger protections: the Violence Against Women Act and the Family Violence Prevention and Services Act prohibit victim service providers from entering survivor information into HMIS at all. Those providers use a “comparable database” that meets the same data and privacy standards but is not shared with the broader community’s HMIS.¹² Survivors are counted in aggregate for the AHAR without being individually identifiable in the shared system.

How the Data Gets Used: System Performance Measures

Beyond raw headcounts, HUD evaluates each CoC on seven System Performance Measures that gauge whether the local response is functioning as a coordinated system. HUD uses these metrics as a competitive factor in the annual CoC Program Competition, so communities with poor performance risk losing funding to those that score better.¹³

The seven measures are:

Length of time people remain homeless.
The share of people who exit to permanent housing and return to homelessness within 6, 12, or 24 months.
Total number of people experiencing homelessness.
Employment and income growth for people in CoC-funded projects.
Number of people who become homeless for the first time.
Homelessness prevention and housing placement for people at imminent risk.
Successful placement from street outreach and successful retention of permanent housing.¹⁴

The data comes from HMIS, which is why clean, complete entry at the local level matters far beyond compliance.

2026 Submission Deadlines

For the 2026 cycle, HUD opened the Homelessness Data Exchange (HDX 2.0) on March 31, 2026, for Continuums of Care to submit Housing Inventory Count and Point-in-Time data. The deadline is April 30, 2026, at 8:00 PM Pacific time.¹⁵ HUD has stated that it does not expect to grant extensions for technical issues encountered on the final day. Meeting this deadline is a scored factor in the annual CoC Program Competition, so a late submission can directly reduce a community’s federal funding. The LSA data feeding Part 2 follows a separate submission timeline, typically later in the year, and CoCs should monitor the HUD Exchange for updated windows.

What the 2024 AHAR Found

The 2024 Part 1 report, based on the January 2024 PIT count, documented approximately 771,480 people experiencing homelessness on a single night, a 10 percent increase from 2023. Sheltered populations rose by 13 percent and unsheltered populations by 7 percent.¹ Roughly 512,000 of those counted were individuals rather than people in families, split almost evenly between sheltered and unsheltered. The figures continue an upward trend, and every gap in a local HMIS record or missed person on a January night can translate into a misallocation of federal resources built on top of these numbers.

1
U.S. Department of Housing and Urban Development. 2024 Annual Homelessness Assessment Report to Congress Part 1 – Point-In-Time Estimates of Homelessness
2
GovInfo. 42 USC 11360a – Annual Homeless Assessment Report
3
Office of the Law Revision Counsel. 42 USC 11313 – Functions
4
HUD Exchange. PIT Count Standards
5
eCFR. 24 CFR 578.7 – Responsibilities of the Continuum of Care
6
HUD Exchange. 2026 HIC and PIT Count Data Submission Guidance
7
HUD Exchange. An Introductory Guide to Submitting Longitudinal Systems Analysis Data for the AHAR
8
HUD Exchange. Four Categories in the Homeless Definition
9
eCFR. 24 CFR Part 578 – Continuum of Care Program
10
HUD Exchange. HMIS Requirements
11
HUD Exchange. FY 2026 HMIS Data Standards Manual
12
HUD Exchange. HMIS Implementation
13
HUD Exchange. System Performance Measures
14
HUD Exchange. System Performance Measures Table Shells
15
HUD Exchange. HDX 2.0 Open for CoCs to Submit 2026 HIC and PIT Count Data