NSDUH stands for the National Survey on Drug Use and Health, an annual survey conducted by the U.S. government that tracks substance use, mental health, and related behavioral health trends across the country. It is the leading source of population-level statistics on alcohol, tobacco, and drug use among Americans aged 12 and older, and it has been running in some form since 1971.
Who Runs NSDUH and Why It Exists
The survey is managed by the Substance Abuse and Mental Health Services Administration (SAMHSA), the federal agency responsible for public health efforts around mental illness, substance use, and addiction treatment. NSDUH exists to give policymakers, researchers, and public health officials a reliable picture of how many people in the U.S. are using various substances, how many are living with mental health conditions, and how those numbers shift from year to year.
The data feeds directly into federal and state decisions about funding, treatment programs, and prevention strategies. When you hear a statistic like “X million Americans had a substance use disorder last year,” that number almost certainly comes from NSDUH.
How the Survey Works
NSDUH is a household survey, meaning interviewers visit people where they live rather than conducting it in a clinic or over the phone. Once a housing unit is selected through a sampling process, an introductory letter goes out, followed by an in-person visit from a trained interviewer.
The interview itself uses two methods. For less sensitive questions (demographics, general health), the interviewer reads questions aloud and records answers on a laptop. For sensitive topics like drug use or mental health symptoms, the survey switches to a system called audio computer-assisted self-interviewing, or ACASI. In this mode, respondents read or listen to questions through headphones and enter their own answers directly on the computer. Nobody else in the room sees what they type. This privacy layer is specifically designed to encourage honest answers about behaviors people might otherwise downplay or hide.
Who Gets Surveyed
The target population is the civilian, non-institutionalized population of the United States aged 12 and older. That broad scope is one of the survey’s strengths, but the “non-institutionalized” part is also one of its most important limitations. People who are incarcerated, living in residential treatment facilities, or experiencing homelessness without a fixed address are generally not captured.
The sample is divided into five age groups: 12 to 17, 18 to 25, 26 to 34, 35 to 49, and 50 or older. Since a 2014 redesign, 25 percent of the sample is allocated to 12- to 17-year-olds, another 25 percent to 18- to 25-year-olds, and the remaining 50 percent to adults 26 and older. This deliberate oversampling of younger age groups gives researchers more precise estimates for the populations where substance use patterns often first emerge.
Sample sizes vary by state. The eight largest states (California, Florida, New York, Texas, Illinois, Michigan, Ohio, and Pennsylvania) historically target around 3,600 completed interviews each, while the remaining states and Washington, D.C., target about 900. Across the entire country, this adds up to roughly 67,000 interviews per year.
What the Survey Measures
NSDUH covers a wide range of topics. On the substance use side, it asks about alcohol, tobacco, marijuana, prescription drug misuse, heroin, cocaine, methamphetamine, hallucinogens, and other drugs. Questions cover not just whether someone has ever used a substance but how recently, how frequently, and whether they meet criteria for a substance use disorder.
On the mental health side, the survey screens for major depressive episodes, serious mental illness, and any mental illness more broadly. It also asks about whether people sought treatment, what barriers they faced, and whether they received the help they needed. This treatment gap data is some of the most cited information NSDUH produces.
The diagnostic criteria behind these measurements have evolved. Substance use disorders were assessed using DSM-IV criteria (an older edition of the standard psychiatric diagnostic manual) through 2019, then shifted to the updated DSM-5 criteria starting in 2020. Major depressive episodes have been measured using DSM-5 standards since 2017. These transitions are worth noting because they can make it tricky to compare numbers directly across certain years.
A Brief History
The survey launched in 1971 under the name National Household Survey on Drug Abuse. In its early decades, it was not conducted every year. Surveys were fielded in 1979, 1982, 1985, and 1988 before becoming annual starting in 1990. In 2002, the name changed to the National Survey on Drug Use and Health to reflect its expanded focus on mental health, not just drug use. That broader scope has made it increasingly central to national conversations about behavioral health.
Known Limitations
Like any self-reported survey, NSDUH is subject to social desirability bias, the tendency for people to underreport behaviors they see as stigmatized (like drug use) and overreport socially approved ones. The ACASI system helps reduce this, but it cannot eliminate it entirely. Research has shown that when participants believe their responses can be independently verified, they report higher rates of drinking and illicit substance use than they do in standard survey conditions.
The exclusion of institutionalized populations is another significant gap. People in jails, prisons, and long-term treatment facilities have substantially higher rates of substance use disorders and mental illness than the general population, so NSDUH’s estimates inherently undercount the full burden of these conditions in the U.S. Homeless individuals without stable housing are also underrepresented. These gaps don’t make the data unreliable, but they do mean the true national numbers are likely higher than what NSDUH reports.
How to Access NSDUH Data
SAMHSA publishes annual national reports with key findings, detailed data tables, and state-level estimates. The most recent release, based on 2023 and 2024 data, is available on SAMHSA’s website. State-level estimates combining 2023 and 2024 survey data are expected by early 2026. Researchers can also access the raw, de-identified datasets for their own analyses through SAMHSA’s data portal, making NSDUH one of the most widely used public health datasets in the country.