Nearly Half of NYC Homeless Deaths Tied to Drugs and Alcohol

August 13, 2026

Among New Yorkers experiencing homelessness, drug-related mortality is the dominant killer, and when alcohol misuse is added, nearly half of all deaths over a recent fiscal year trace back to substances—a stark confirmation of how addiction now defines the risk landscape for this population.

At a Glance

Drug-related deaths were the leading cause of death among New Yorkers experiencing homelessness, accounting for 40% of all deaths in FY25.
Alcohol misuse/dependence added another 5%, bringing substance-involved causes near the halfway mark of total homeless deaths.
The city identified 634 deaths among people experiencing homelessness in FY25; drug-related mortality declined in count from the prior year yet remained predominant.
This pattern—overdose and drug-related deaths atop the mortality table—matches a long-running New York City trend documented over multiple reporting cycles.

What the latest surveillance shows

New York City’s twentieth annual mortality report for persons experiencing homelessness is unambiguous on its core finding: drug-related causes led all categories, comprising 40% of deaths in Fiscal Year 2025, followed by heart disease, accidents not involving drugs, alcohol misuse/dependence, and cancer. The city recorded 634 deaths for the period July 1, 2024 through June 30, 2025; the number of drug-related deaths was 251, and alcohol misuse/dependence accounted for an additional 34, together summing to 45% of all deaths. The report also notes that drug-related deaths fell in count compared with FY24—a decline from 348 (45%) to 251 (40%)—yet the category remained the leading cause, consistent with citywide and national overdose patterns.

Those proportions sit squarely within the city’s broader public-health picture. New York City’s overdose data for the general population show a large, fentanyl-driven epidemic, with opioids involved in the bulk of fatal overdoses and substantial year-over-year variation as prevention, supply, and treatment dynamics shift. Within the homeless population, where comorbidities such as mental illness, chronic disease, and trauma are more prevalent, the same overdose engine registers with even greater force—an effect surveillance has documented repeatedly over the past decade.

How the numbers are built: classifications, sources, and scope

New York City’s homeless-deaths surveillance program integrates case ascertainment from the Department of Social Services and the Office of Chief Medical Examiner, then assigns underlying cause-of-death using standardized codes. “Drug-related” is a defined category; “alcohol misuse/dependence” is a separate one. Aggregating them illuminates the total substance burden, but the two are not interchangeable, and the report treats them distinctly for analytic clarity. This taxonomy—built on death certificates and medical examiner determinations—has been in place in some form since the city’s early homeless mortality studies and has supported consistent, longitudinal comparisons.

The FY25 report’s top-line picture is methodologically straightforward: a known denominator of 634 deaths, cause categories with counts and shares, and an explicit year-over-year comparison for drug-related mortality. Like all cause-of-death surveillance, it carries boundary questions (for example, polysubstance involvement or indeterminate intent), and a small subset remains pending or undetermined at any point in time. These caveats do not alter the central fact pattern: drug-related causes dominate the table, and alcohol-related deaths sit within the top five.

What history tells us: a through-line from the first reports to today

If the FY25 numbers feel grimly familiar, it is because they are. The city’s earliest sustained work on homeless mortality—launched in the mid-2000s—identified drug overdose as a leading cause among single adults in shelter, with shares in the range of a quarter to a third of deaths in initial years. Subsequent analyses, spanning sheltered and unsheltered groups, found overdose remained at or near the top, even as other causes like heart disease, cancer, and accidents retained significant shares. Over time, fentanyl’s diffusion amplified the risk profile; surveillance captured that escalation, and in recent cycles, overdose and drug-related categories consolidated their lead position across settings.

This continuity matters. It means FY25 is not an anomaly to be rationalized away; it is the latest datapoint in a long arc that has seen substance-involved mortality—particularly involving opioids—become the signature lethal threat among people experiencing homelessness in New York City. Independent reviews of local homeless mortality efforts have pushed for standardization across jurisdictions; New York’s long-run consistency is precisely the sort of infrastructure those reviews aim to institutionalize.

Interpreting “nearly half”: accuracy without oversimplification

“Nearly half” is a fair short-hand only when readers understand it combines two separate categories: drug-related at 40% and alcohol misuse/dependence at 5% in FY25. Keeping the categories distinct guards against the common analytical mistake of collapsing alcohol-specific pathologies—alcoholic liver disease, withdrawal syndromes—into overdose dynamics and vice versa. Both are substance-involved; both are preventable; each requires different levers of response. Public accountability relies on precision here: overdose prevention and naloxone saturation address acute drug poisoning; alcohol dependence treatment and hepatology pathways target chronic alcohol morbidity and mortality.

The discipline of clear definitions also keeps other causes properly in view. Heart disease, injuries excluding overdoses, cancer, exposure, and homicide collectively account for the majority of deaths not attributed to substances. Policy that chases only overdoses without addressing cardiovascular care, injury prevention, and safe shelter environments will not bend the overall mortality curve as far as it could.

This study sheds crucial light on the urgent need to address substance abuse as a major factor contributing to homeless deaths in NYC, and it's commendable that researchers are working tirelessly to uncover the root causes of this complex issue and find potential solutions to…

— J. LEE (@JWPAX3051) August 13, 2026

Mechanisms and leverage points: why substance mortality concentrates among people experiencing homelessness

Three mechanisms explain the outsized substance mortality burden in this population. First, risk environment: unsheltered and congregate settings elevate exposure to potent street drug supplies and impede consistent harm-reduction practices. Second, comorbidity and care discontinuity: unmanaged mental illness, infectious diseases, and chronic conditions complicate use patterns and raise fatality risk; fragmented coverage and episodic access interrupt medication-assisted treatment and withdrawal management. Third, pharmacologic potency: fentanyl and analogs compress the margin for error, making brief lapses in tolerance—after detox, incarceration, or hospitalization—especially dangerous.

Interventions map to these mechanisms. Harm reduction—widespread naloxone carriage, drug-checking where permitted, overdose-prevention education—targets the immediate poisoning risk. Treatment-on-demand for opioid use disorder with buprenorphine or methadone reduces all-cause and overdose mortality. Low-threshold housing with integrated primary care improves management of chronic disease, while assertive outreach and post-hospital “warm handoffs” reduce fatal gaps in care. Cities that link shelter and supportive housing to on-site addiction and medical services consistently report better retention and lower mortality than those that silo functions.

Trends and signals: decline in drug-related counts, but not the threat

The FY25 report documents a 28% decline in drug-related deaths compared with FY24, from 348 to 251, even as drug-related causes remained the leading category. Citywide overdose surveillance in 2024 also reported a substantial decrease in deaths compared with 2023, indicating a broader movement in the same direction—likely reflecting combined effects of harm reduction, treatment access, and potentially shifting supply characteristics. The implication is not that the crisis has passed; it is that policy can move the needle when it aligns with the pharmacology and the lived realities of use. Sustained effort determines whether a single-year decline hardens into a trend.

This is where precision in messaging matters. Overstating “nearly half” risks making other lethal causes invisible; understating the role of drugs and alcohol blunts urgency where rapid, tactical interventions save lives. The correct framing is clinical and practical: substance-involved causes are the leading killers among New Yorkers experiencing homelessness; other causes form a substantial remainder; both sets require tailored, concurrent action.

What this means for policy and practice

For city agencies and provider networks, the FY25 profile sets a clear operational agenda. Keep overdose prevention at saturation levels in shelters, supportive housing, street outreach, and emergency departments; ensure medication for opioid use disorder is available without delay or paperwork gauntlets; integrate alcohol-dependence care and hepatology consults into homeless health services; and treat discharge from detox, jail, and hospitals as red-alert windows for fatal relapse. Simultaneously, invest in cardiovascular risk management and injury prevention within homeless services, because a mortality table is an instruction manual as much as it is a ledger.

Finally, accountability depends on maintaining—and improving—the measurement system. Publishing full methodology, coding notes, and subgroup tables allows independent replication and sharper targeting. New York City has already built one of the longest-running homeless-mortality surveillance programs in the country; continuing to open its technical spine ensures the public debate remains anchored in evidence rather than rhetoric.