Skip to main content

Data Source

State Data Tables and Reports From the 2019-2020 NSDUH

Find reports that present findings and data for U.S. states based on results of the combined 2019 and 2020 National Survey on Drug Use and Health (NSDUH).

NSDUH State Result Data Tables

NSDUH is an annual survey of the U.S. civilian, noninstitutionalized population ages 12 years or older. The combined 2019 and 2020 NSDUH reports provide state estimates for select measures of substance use and mental health outcomes by age group (e.g., ages 12 or older, 12 to 17, 18 to 25, and 26 or older and 12 to 20 for alcohol measures). State and regional estimates are based on a small area estimation (SAE) methodology in which state-level NSDUH data are combined with county and sub-county level census data from the state.

The 2019-2020 NSDUH State Prevalence Estimates (PDF, Excel) provide prevalence estimates and 95% confidence intervals by age group. CSV files can also be downloaded.

The 2019-2020 NSDUH Estimated Totals By State (PDF, Excel) provides the estimated totals and 95% confidence intervals by age group. CSV files can also be downloaded.

2020 NSDUH Frequently Asked Questions

  1. Why does SAMHSA caution against comparing 2020 estimates with prior years’ estimates?
    • The coronavirus disease 2019 (COVID-19) pandemic necessitated methodological changes in data collection. The COVID-19 pandemic could have affected both the true prevalence of a behavior (or mental health issue) and the behavior’s measurement through the necessitated methodological changes. The primary methodological changes were (1) virtually no data collection from mid-March through September 2020, (2) introduction of web data collection in October 2020 with very limited in-person data collection, and (3) questionnaire changes beginning in October 2020. These changes could result in data collection mode effects or other effects on the estimates. Because these changes in data collection coincided with the spread of the COVID-19 pandemic and any related behavioral or mental health changes, we cannot fully separate the effects of methodological changes from true changes in the outcomes.
    • Cautioning against comparisons does not necessarily mean the data are not comparable. For most estimates, comparability is unknown.
  2. What impact did adding a new data collection mode (web) have on the estimates?
    • Because the introduction of the web mode to NSDUH coincided with the effects of the COVID-19 pandemic on people’s behavior and mental health, it currently is not possible to quantify the effects of the web mode on the estimates.
  3. If SAMHSA is asking us to exercise caution when comparing 2020 data with prior years, why then can the 2020 data be combined with prior year data to generate the state-level estimates?
    • At issue here are two separate but seemingly similar concepts; comparing and combining. We ask that people use caution when comparing 2020 with any single prior year since there were changes in how we collected the data (i.e. going from solely in-person interviewing to a combination of web-based and in-person interviewing). For example, suppose we see an apparent increase from 2019 to 2020 in the percentage of major depressive episodes (MDEs) in a particular group. It may be that the difference we are seeing reflects an actual change in that population regarding MDE. Or it may be that simply asking the questions by a different method (e.g. web vs in-person) influenced the way respondents answered the depression questions, and that there is no actual increase in MDE in that group. At this point, there is no way for us to tell which of these things happened. Therefore, we ask that people use caution when making that comparison.
    • But when we are talking about combining the data, that is a different concept. Combining data from multiple years is necessary to provide enough data to calculate precise estimates at the state and sub-state level. So, back to our MDE example, the methods we used to collect that data have changed, but the scales we use to assess MDE have not. Also, since we are interested in the average across those years, any effect that the method change may have had will be lessened to a negligible level, and the resulting estimate will still be valid.
    • One measure, i.e. substance use disorder, not only was subject to the same methodology change as described above, but also was subject to a change in the scales that we use to assess them (going from DSM-IV criteria to DSM-5 criteria). Because of this change, it would not be valid to combine the data to calculate estimates for substance use disorder (or any other estimates that are reliant on substance use disorder). And in fact, these estimates were not calculated and will not be part of the state and sub-state report.

      Please direct any questions regarding the data included in this report to

      Read the entire 2020 NSDUH report, highlights from it or frequently asked questions about it. People searching for treatment for mental or substance use disorders can find treatment by visiting or by calling SAMHSA’s National Helpline, 1-800-662-HELP (4357) .Reporters with questions should send inquiries to

NSDUH State Result Maps

The survey also includes maps for certain estimates and detailed tables including percentages and counts for each state, census region, and the nation, by age group.

The 2019-2020 NSDUH National Maps of Prevalence Estimates by State (PDF, HTML) show 2019-2020 estimates, sorted from lowest to highest for each measure, and divided into five categories. More information about quintiles used is available (PDF, HTML).

State and Regional Methodology Reports

State and regional estimates are based on a small area estimation methodology in which state-level NSDUH data are combined with county and census block group/tract-level data from the state.

The 2019-2020 NSDUH: Guide to State Tables and Summary of Small Area Estimation Methodology Report (PDF, HTML) is a guide to the development and presentation of 2019-2020 state estimates. A summary of the methodology is included as well as sample sizes, response rates, and population estimates.

The 2019-2020 NSDUH: Other Sources of State-Level Data report provides a brief description of the Behavioral Risk Factor Surveillance System (BRFSS) and presents a comparison of state-level BRFSS and NSDUH prevalence estimates for alcohol and cigarette use among people ages 18 or older.

State versus other states and regional comparisons are available as p-value tables in a zipped file of CSV files and Excel formats. A Guide to P-Value Tables is also available.