If you just saw a headline saying heroin admissions fell after marijuana became legal for adults, you’re asking the right question. A 2026 analysis from Penn State linked the two later stages of legalization with a lower rate of heroin involvement among substance use treatment admissions. That is an association in admissions data, not proof that legalization lowers heroin use. Below, you’ll find how the study was built, its numbers, the results that didn’t reach statistical significance, and where to turn if opioid use is a worry for you.
Affiliate disclosure: You will find affiliate links on this page. This site may receive a commission on purchases made through them.

Quick facts, as of 2026-11-20
| Item | Detail |
|---|---|
| Study | Heroin involvement in U.S. substance use treatment admissions, by state cannabis policy stage |
| Published | Online 11 November 2026 in the International Journal of Drug Policy, 2026;158 |
| Team | Three authors at Penn State’s Ross and Carol Nese College of Nursing |
| Data | Treatment Episode Data Set-Admissions (TEDS-A), 1992 to 2023. SAMHSA’s Treatment Episode Data Set covers admissions to and discharges from substance use treatment services, as reported by state agencies. |
| Measure | Rate per 1,000 treatment admissions |
| Main heroin result | Fewer heroin-involved admissions in the two later stages, statistically significant under the primary weighting |
| Headline coverage | NORML, an advocacy group for legalization, 19 November 2026 |
What did the Penn State study find?
The study compared how often heroin showed up in treatment admissions across states at five stages of cannabis policy. The two later stages, adult-use legal without retail and operational adult-use retail, were linked with fewer heroin-involved admissions. Compared with states that had no coded cannabis policy, the estimates range from 102.81 to 127.71 fewer heroin-involved admissions per 1,000 treatment admissions, depending on the stage and on how heroin was counted.
Who ran the study and where was it published?
Three authors at Penn State’s Ross and Carol Nese College of Nursing wrote the paper. It was published online on 11 November 2026 in the International Journal of Drug Policy. You can check the citation on the PubMed record for the paper.
How the study compared states
The design is called difference-in-differences. In plain terms, it compares how heroin admissions changed in states that adopted a cannabis policy with how they changed in states that did not. States adopted policies at different times, so the study uses a staggered-adoption version of that method. The data are TEDS-A records from 1992 to 2023, and each state-year was coded into one of five policy stages. The authors also ran checks using event-time and equal-state-year weighting.
The five policy stages
- None: no coded cannabis policy. This is the comparison group.
- Medical authorization without sales.
- Operational medical access.
- Adult-use legal without retail.
- Operational adult-use retail.
Stages 2 and 3 are medical. Stages 4 and 5 are adult-use.
The numbers, per 1,000 treatment admissions
| Stage | Outcome | Estimate per 1,000 admissions | 95% confidence interval |
|---|---|---|---|
| Adult-use legal without retail | Any heroin involvement | -107.01 | -190.91 to -23.12 |
| Operational adult-use retail | Any heroin involvement | -127.71 | -225.80 to -29.63 |
| Adult-use legal without retail | Primary heroin | -102.81 | -182.39 to -23.24 |
| Operational adult-use retail | Primary heroin | -126.45 | -221.47 to -31.44 |
Read -107.01 as about 107 fewer heroin-involved admissions for every 1,000 treatment admissions. These are rates per 1,000 admissions, not percentages of people. The range in brackets is the 95% confidence interval, the span where the true figure most likely falls.
What didn’t reach statistical significance?
Two results did not reach statistical significance. The medical-policy stages did not, and neither did the outcome for other opiates and synthetics.
The heroin result also depends on how states are weighted. When each state-year got equal weight, the later-stage heroin estimates stayed negative but were smaller, and they were not statistically significant. So the heroin result is significant under the primary weighting, and not under the alternative one.
What a treatment admission measures, and who it counts
A treatment admission is one entry into a substance use treatment program. This study measures how often heroin was involved in those admissions. It does not measure how many people use heroin in the population, and it does not count overdoses. TEDS counts admissions, not people. The people counted are those admitted to substance use treatment in the U.S.
Where the headline came from
NORML, an advocacy group for cannabis legalization, ran an item on 19 November 2026 under the headline “Analysis: substance abuse admissions for heroin fall following adoption of adult-use marijuana legalization.” That headline reads as if legalization caused a fall in heroin use. The paper reports an association in treatment-admission rates, and that result is significant only under the primary weighting. You can read the NORML news item yourself.
Common mix-ups
- Admissions versus use. The paper counts treatment admissions. It does not count how many people use heroin.
- Timing versus cause. A headline that says admissions “fall following” legalization describes timing. The paper reports an association.
- Heroin versus all opioids. The other-opiates-and-synthetics outcome did not reach statistical significance, so the heroin result doesn’t cover every opioid.
- Medical versus adult-use. The medical-policy stages did not reach statistical significance. The two adult-use stages did, under the primary weighting.
- Overdoses. This study does not measure overdoses.
An observational study like this shows an association, not proof that legalization lowered heroin use.
Where to turn if opioid use is a concern for you
If opioid use, including heroin, is a concern for you or someone close to you, a doctor is the right first call. A pharmacist can also help you find the next step. Related on this site: our guide to cannabis use disorder: what it is and treatment and our cannabis and mental health guide.
Questions that settle your next step
- Am I reading about use, or about treatment admissions? It’s admissions.
- Does the heroin result hold under both weightings? It is significant under the primary weighting only.
- Does the headline match the paper? Compare NORML’s wording with the PubMed record for the paper.
By the Global MH Summit Research Team
This article is for general information purposes only and does not constitute medical advice. Consult your doctor or qualified healthcare provider before making changes to your health routine.