Key Takeaways
- A study reveals that cannabis legalization, particularly retail access, correlates with lower heroin involvement in treatment admissions.
- Researchers analyzed over 55 million admission records from SAMHSA’s Treatment Episode Data Set, focusing on distinct cannabis policy stages.
- The study found 107 fewer heroin-related admissions per 1,000 during adult-use legalization and 128 fewer once retail sales began.
- Retail access to cannabis appears to have a stronger impact on reducing heroin treatment admissions than legalization alone.
- While the findings provide valuable insights, researchers caution against overextending conclusions about overall heroin use in the population.
Cannabis policy research keeps growing, and another entry adds something useful to the conversation around legalization and opioids. Published in the International Journal of Drug Policy, the study pulled data from nearly every substance use treatment admission reported in the country over three decades. The researchers wanted to know if legal cannabis access, and specifically the opening of retail stores, lines up with changes in how often heroin shows up in those admission records.
This isn’t the first research we have covered connecting cannabis access to lower opioid-related harm. We’ve reported on studies tied to opioid overdoses, opioid prescriptions, and even unregulated opioid use. This new study adds a different angle: treatment admissions, tracked using one of the most detailed government data sets in the country.
What Did the Study Actually Look At?
Researchers analyzed roughly 55.2 million admission records from SAMHSA’s Treatment Episode Data Set, commonly known as TEDS-A. This data set tracks admissions to licensed treatment facilities across all 50 states from 1992 through 2023.
Instead of lumping every form of cannabis legalization together, the team split state policy into five distinct stages. These included having no cannabis policy at all, authorizing medical cannabis without active sales, having fully operational medical dispensaries, legalizing adult-use cannabis before retail stores opened, and finally, operational adult-use retail sales. That level of detail matters, because legalizing something on paper and actually being able to walk into a store and buy it are two very different realities.
The study used a statistical method called staggered-adoption difference-in-differences, which compares what actually happened in legal states against a modeled prediction of what might have happened had those states never legalized. Read the full abstract and methodology here: doi.org/10.1016/j.drugpo.2026.105593.
What Were the Key Findings on Heroin?
The headline numbers focus on heroin involvement in admissions. According to the study, states with adult-use legalization but no retail stores yet open saw an estimated 107 fewer heroin-involved admissions per 1,000 reported admissions, compared to the modeled counterfactual. Once retail sales became operational, that gap widened to an estimated 128 fewer admissions per 1,000.
The pattern held when researchers looked specifically at heroin listed as the primary substance behind the admission, with similar reductions reported at both policy stages. The authors describe this as evidence that both legalization and the launch of retail access line up with a lower share of heroin involvement in treatment data.
It’s worth pointing out that this research treats legalization as a layered, ongoing process rather than a single event. Passing a law and opening a dispensary down the street aren’t the same milestone, and the data seems to back that distinction up.
Why Does the Retail Stage Matter More Than Legalization Alone?
One of the more interesting parts of this study is how it separates legal status from actual access. A state can legalize adult-use cannabis and still take a year or more before stores open their doors. That gap matters for research like this, because people can’t act on access they don’t yet have.
The study found that once retail sales began, the association with lower heroin involvement grew stronger. This lines up with a broader theme we’ve seen across other research Beard Bros has covered, including studies tied to dispensary openings and opioid prescriptions. Real, physical access to cannabis products seems to carry more weight than legal status alone.
Should This Be Read as Proof Cannabis Reduces Heroin Use?
Here’s where some honesty matters. The researchers are careful to note that their alternative statistical method, which weighted each state year equally, produced smaller estimates that weren’t statistically significant. That doesn’t erase the primary findings, but it does mean the results aren’t bulletproof.
It’s also worth separating two different ideas. This study measures the share of treatment admissions that involve heroin, not the total number of people using heroin across the entire population. A shrinking share of admissions could reflect several things happening at once, including shifts in who seeks treatment, changes in how substances get reported, or trends in opioid use outside of cannabis policy entirely. The researchers themselves describe the evidence linking cannabis policy to opioid outcomes as inconsistent across different studies and methods, which is a fair and useful reminder.
None of this discredits the findings. It just means this is one solid data point in a growing, evolving research conversation, not a finished verdict.
How Does This Fit With Other Cannabis and Opioid Research?
This study doesn’t exist in a vacuum. We have tracked a handful of studies over the past few years pointing in a similar direction. A University of Kentucky study published in Preventive Medicine Reports looked at more than 107 million commercially insured adults and found notable drops in non-fatal opioid poisonings tied to both medical dispensaries and recreational cannabis laws. You can read our coverage of that research here.
Separately, University of Georgia researchers found that opioid prescription rates dropped meaningfully in states with legal medical cannabis, a trend we covered in this article. And on a more personal level, a federally funded survey found that over half of participants who used both cannabis and unregulated opioids reported reduced opioid use during periods of active cannabis use, which we wrote about here.
Taken together, these studies build a pattern worth paying attention to, even if no single one closes the book on the question.
What This Study Adds to the Conversation
What makes this particular study stand out is its scale and its structure. Few data sets track 55.2 million admissions across every state over 31 years. And the decision to separate legal authorization from retail access gives researchers (and readers) a clearer picture of what’s actually driving the trend.
If you’re following the long-running debate over cannabis and opioids, this is a study worth bookmarking. It won’t settle every argument, but it adds real weight to the growing body of evidence suggesting legal cannabis access may play a meaningful role in how people interact with opioids like heroin. For more ongoing coverage of cannabis research and policy, check out our news section or sign up for The Friday Sesh to get it delivered straight to your inbox.
Frequently Asked Questions
The study measured heroin involvement in substance use treatment admissions across five cannabis policy stages, using TEDS-A data from 1992 to 2023, and compared outcomes to a modeled no-legalization scenario.
The study found a lower share of heroin-involved treatment admissions associated with adult-use legalization and retail sales, but it doesn’t prove a population-wide drop in heroin use.
TEDS is a national data system run by SAMHSA that tracks admissions to licensed substance use treatment facilities across the United States.
The study estimated 107 fewer heroin-involved admissions per 1,000 during adult-use legalization, and 128 fewer per 1,000 once retail sales opened.
Yes. An alternative weighting method produced smaller, non-significant results, and the study measures admission share, not total heroin use across the population.