US Drug Overdose Deaths (CDC provisional, monthly)
Provisional counts of drug overdose deaths in the United States, for the US as a whole and each state (plus Washington DC, New York City and Puerto Rico), from CDC's National Center for Health Statistics (NCHS). Each value is the number of deaths in the 12 months ending in the given month, from 12 months to January 2015 onward. For each jurisdiction and month there are counts of all overdose deaths and of deaths involving the main drug classes (synthetic opioids such as fentanyl, psychostimulants such as methamphetamine, cocaine, heroin, prescription-type opioids, methadone), both as reported and as predicted by NCHS after adjusting for delayed reporting. Recent months are provisional and will rise as records arrive. Updated monthly.
Overdose deaths in the 12 months ending each month, by jurisdiction and drug class
Field
Type
Description
Constraints
date
date
Last day of the month that ends the 12-month period. 2026-06-30 means deaths from 1 July 2025 to 30 June 2026.
jurisdiction
string
Two-letter code: US (the 50 states and DC; excludes Puerto Rico), a state, DC, YC (New York City) or PR (Puerto Rico). NY excludes New York City.
jurisdiction_name
string
Name of the jurisdiction.
indicator
string
What is counted: all_drugs (all drug overdose deaths), opioids (any opioid, T40.0-T40.4, T40.6), synthetic_opioids (synthetic opioids other than methadone, mainly fentanyl, T40.4), psychostimulants (psychostimulants with abuse potential, mainly methamphetamine, T43.6), cocaine (T40.5), heroin (T40.1), natural_semisynthetic_opioids (natural and semi-synthetic opioids such as oxycodone, T40.2), methadone (T40.3). A death can involve several drugs, so drug classes overlap and do not sum to all_drugs.
Reported provisional count of deaths. Empty when CDC suppresses it (fewer than 10 deaths) or does not show it (low data quality).
predicted_deaths
integer
NCHS predicted count, adjusted for delayed reporting. Never below the reported count in this data; empty where CDC does not publish one.
note
string
suppressed (fewer than 10 deaths), low_quality (not shown by CDC due to low data quality), or empty.
enum ["","suppressed","low_quality"]
incomplete
boolean
true where CDC flags the count as underreported due to incomplete data (the most recent months).
Reporting completeness by jurisdiction and month
Field
Type
Description
date
date
Last day of the month that ends the 12-month period.
jurisdiction
string
Two-letter jurisdiction code, as in overdose-deaths.
jurisdiction_name
string
Name of the jurisdiction.
all_cause_deaths
integer
Provisional count of deaths from all causes in the 12-month period.
percent_complete
number
Percentage of expected death records received (CDC's completeness measure for the period).
percent_pending_investigation
number
Percentage of death records whose cause is still pending investigation.
percent_with_drugs_specified
number
Percentage of drug overdose deaths with at least one specific drug named on the death certificate. Low values make drug-class counts less reliable.
Monthly counts of drug overdose deaths in the United States, for the country as a whole and for each state, Washington DC, New York City and Puerto Rico, from the CDC's National Center for Health Statistics (NCHS). Each value is the number of deaths in the 12 months ending in a given month, from the 12 months to January 2015 onward (CDC's 10 November 2026 release ends at June 2026). As well as all overdose deaths, there are counts for the main drug classes: synthetic opioids (mainly fentanyl), psychostimulants (mainly methamphetamine), cocaine, heroin, natural and semi-synthetic opioids (such as oxycodone), methadone and all opioids. Units are deaths.
Recent figures are provisional. Death records arrive late and causes can be pending investigation, so the latest months are undercounted and will rise. NCHS publishes both the reported count and a predicted count that adjusts for this delay.
What stands out
In CDC's 10 November 2026 release, 64,433 overdose deaths were reported in the 12 months to June 2026, about 42% below the peak of 111,466 in the 12 months to June 2023. That is the lowest 12-month total since the 12 months to December 2016 (63,938). The latest figure is provisional; adjusted for delayed reporting, NCHS predicts 65,724, which is 41.5% below the predicted peak of 112,418 (12 months to August 2023).
Data
overdose-deaths.csv
One row per jurisdiction, month and indicator.
date: the last day of the month that ends the 12-month period. 2026-06-30 means deaths from 1 July 2025 to 30 June 2026. Values for consecutive months overlap by 11 months, so they show a rolling annual total, not monthly deaths.
jurisdiction, jurisdiction_name: US (the 50 states and DC; Puerto Rico is not included), each state, DC, YC (New York City) and PR (Puerto Rico). CDC reports New York City separately, so NY is New York State excluding New York City.
indicator: all_drugs (all drug overdose deaths), opioids (any opioid), synthetic_opioids (synthetic opioids other than methadone, mainly fentanyl), psychostimulants (psychostimulants with abuse potential, mainly methamphetamine), cocaine, heroin, natural_semisynthetic_opioids (such as oxycodone and hydrocodone) and methadone. One death can involve several drugs, so the drug classes overlap and do not add up to all_drugs.
reported_deaths: the provisional count of deaths reported so far.
predicted_deaths: the NCHS count adjusted for delayed reporting. It is never below the reported count; the gap is largest for the most recent months.
note: suppressed where CDC hides counts under 10, low_quality where CDC does not show the drug-specific count because of low data quality; both leave reported_deaths empty.
incomplete: true where CDC flags the count as underreported due to incomplete data (the most recent six months; January to June 2026 in the 10 November 2026 release).
reporting-quality.csv
One row per jurisdiction and month, with CDC's measures of how complete the provisional data is: all_cause_deaths (deaths from all causes in the 12 months), percent_complete (share of expected death records received), percent_pending_investigation (share of deaths whose cause is still pending) and percent_with_drugs_specified (share of overdose deaths that name at least one specific drug).
Louisiana has only all-drug counts; CDC does not publish its drug-class counts.
Caveats: reporting of specific drugs varies between states, so CDC advises against comparing drug-specific counts across jurisdictions. Each monthly release revises earlier months, so figures here may differ from those quoted elsewhere. Final annual data from NCHS may differ from these provisional counts.
Licence: the data.cdc.gov dataset page gives its licence as "Public Domain U.S. Government". It is a work of the US federal government. Suggested citation from CDC: National Center for Health Statistics. VSRR Provisional Drug Overdose Death Counts.
Updated monthly by a scheduled job, shortly after CDC's mid-month release. First published from CDC's 10 November 2026 release (fetched 2026-11-22); the monthly job refreshes data/ and archive/, so see archive/ for the current release.
API access
Access dataset files directly from scripts, code, or AI agents.
Each file has a stable URL (r-link) that you can use directly in scripts, apps, or AI agents. These URLs are permanent and safe to hardcode.