Minnesota studied 6,621 cancer patients on medical cannabis. Here is what they actually reported.

By Jaycub | MN Cannabis Hub | Published September 16, 2026

On September 11 the Office of Cannabis Management put out a news release about cancer patients in the state medical program. The useful document is the 32-page report behind it: Impact of Medical Cannabis on Minnesota Cancer Patients, dated September 10, 2026.

I read the PDF. The release rounded. The tables did not.

Between July 1, 2015 and December 31, 2023, 7,762 people enrolled in Minnesota's medical cannabis program with a cancer-related qualifying condition. 6,621 of them, 85.3 percent, actually bought product. The report is about the buyers.

That is a large Minnesota file. It is also observational. Patients rated their own symptoms before each purchase. Nobody was randomized. Nobody was given a placebo. If you came here looking for a miracle claim, close the tab. If you came here because you, or someone you love, is trying to decide whether a Minnesota medical card is worth the paperwork for cancer symptoms, stay.

What Minnesota actually measured

Minnesota's medical program has been required, since it started, to study the people in it. That is unusual. Most state programs hand out cards and stop. Ours built a registry that captures what patients, certifying clinicians, and manufacturers report, then OCM's Division of Medical Cannabis pulls a de-identified research set.

Cancer has been a qualifying condition since day one, July 1, 2015, under three symptom gates:

• cancer with severe or chronic pain
• cancer with nausea or severe vomiting
• cancer with cachexia or severe wasting

Patients can qualify under more than one. In this cohort, 69.2 percent (4,581) qualified on pain, 49.3 percent (3,264) on nausea or vomiting, and 29.8 percent (1,973) on cachexia or wasting. OCM does not collect cancer type or stage. That is a hole in the file, and the report says so.

Before each purchase, patients fill out a Patient Self-Evaluation. They rate eight symptoms over the last 24 hours on a 0 to 10 scale: anxiety, lack of appetite, depression, disturbed sleep, fatigue, nausea, pain, and vomiting. OCM treats a 30 percent or greater drop from that person's own baseline as a clinically significant change. A four on the scale is the floor for "moderate to severe."

The four-month window is the one the press release used. The report also tracks who still had that improvement four months later. Those two numbers are not the same thing, and the second one is the one a lot of coverage skipped.

The numbers, in one table

Table 4.1A in the report covers 6,605 of the 6,621 purchasers (a handful did not have usable symptom scores). Among people who started with moderate to severe symptoms:

Source: OCM, Table 4.1A, September 10, 2026.

Read the table left to right. Vomiting is the headline in the release because 47.2 percent of people who started moderate-to-severe reported a 30 percent drop inside four months. Pain is the condition most patients qualified on, and it is also one of the weaker four-month results: 29.9 percent. Fatigue is everywhere at baseline (91.5 percent) and barely moves on the "got better and kept it" column (10.8 percent).

The last column is the honest one. A lot of people leave the program. OCM says that out loud. Chemotherapy windows are often shorter than eight months. Some patients are already in hospice when they enroll. Minnesota does not collect data on other medical treatments, so the report cannot tell you who dropped out because the cannabis did not help and who dropped out because treatment ended.

That is why the "a quarter of patients reported and maintained an improvement" line in the release exists. It is a different statistic from the four-month snapshot. Both are in the PDF. Use both.

If pain is the reason you are here, our cannabis for pain in Minnesota guide is the practical sibling to this study. If sleep is the wrecked piece, start with cannabis for sleep.

What patients bought

In each person's first year in the program, OCM counted 35,859 sales and 121,865 products across those 6,621 patients.

Inhalation (flower and vape) was 48 percent of products. Enteral (gummies and capsules) was 42.5 percent. Among inhalation products, high to very high THC:CBD ratios were 58.9 percent of what got bought. Among edibles and capsules, that high-THC share was 82.4 percent.

Flower that is sold as raw flower is not classified the same way. OCM notes that those SKUs do not carry milligram THC/CBD the way other products do, so they get a cruder high/low split by percent of weight.

None of that is a shopping list. It is a record of what this cohort actually paid for. If you are building a first medical run, talk to a clinician about dose. The report does not pick a product for you.

OCM's September 11 release quotes Dr. Dylan Zylla, an oncologist and medical director of the HealthPartners Cancer Research Center: "I recommend starting with a low dose of THC and adjusting up gradually every few days to reduce the risk of side effects." That is his quoted advice in the release, not a Hub protocol. The report itself tells readers to talk to a medical professional about dosage before starting.

Who these patients are

Average age at enrollment: 58. Range: 1 to 100. The overall medical program averages 47, so this cancer group is older, which matches the disease.

Gender is basically even: 50.4 percent female (3,334), 49.1 percent male (3,251), 0.5 percent prefer not to answer.

Race: 88.0 percent white (5,828), 3.2 percent Black, African, or African American (213), 1.7 percent Asian (112), 1.4 percent two or more races (90), 1.1 percent American Indian or Alaska Native (72). OCM notes that 88 percent white is higher than the overall medical program (79.8 percent) and higher than the Census figure it cites for Minnesota (82.6 percent). The cohort is not a mirror of the state. It is a mirror of who enrolled with a cancer qualification and bought product.

Geography is the part that made this feel like Minnesota instead of a journal abstract. OCM mapped patients by the first three digits of ZIP:

Source: OCM, Table 2.6.

Almost three in four of these patients live in the Twin Cities ZIP regions. Rochester, sitting next to Mayo, is 3.9 percent of the file. Outstate is in the file. It is a smaller share.

Side effects, as reported

About 15 percent of patients reported an adverse side effect on a self-evaluation. Of those reports, 60 percent were mild, 34 percent moderate, and 6.2 percent severe. Dry mouth was the most common side effect overall. Fatigue was the most common severe one.

The registry has a blind spot here too. If someone had a bad reaction and never bought again, the last side effect is not in the file. OCM says that.

What this does not prove

The report cites the research that does not go the way advocacy decks go. A 2023 review (Häuser et al.) found moderate-certainty evidence that nabiximols and THC were not more effective than placebo for opioid-refractory cancer pain. A 2020 meta-analysis (Boland et al.) found no difference in pain scores versus placebo. Appetite and cachexia reviews (Mücke 2018, Simon 2022) found little evidence cannabinoids beat placebo on calories or weight.

Then it puts Minnesota's own earlier work next to that: Anderson et al. 2019, on this same program from 2015 to 2017, and a 2021 Minnesota pilot RCT (Zylla et al.) in Stage IV patients that did see a signal on opioid reduction when medical cannabis started earlier. Those citations are in the PDF. They are not a dunk on the new tables. They are the reason the new tables have to stay modest.

This file tells you what thousands of Minnesota patients said happened to their scores after they started buying. It does not tell you cannabis caused the change. It does not tell you what will happen for any one person. It does not replace an oncologist.

Dr. Grace Christensen, an OCM senior research analyst, said in the September 11 release: "These results show great promise in using cannabis for treating the symptoms that occur with cancer and cancer-related treatments. Even minor relief can improve patients' quality of life and help them better manage battling the disease." Promise is the right word. Proof of causation is not in this dataset.

OCM's release also says the medical program has more than 90,000 registered patients, that enrollment doubled after the Legislature eliminated patient fees in 2023 and expanded qualifying conditions in 2024, and that most states see medical enrollment fall after adult-use legalization while Minnesota's rose. Those three claims are OCM's, from the release. They are not in the cancer PDF's tables, and we have not independently verified the national comparison.

If you or someone you love has cancer

Cancer is already a qualifying condition in Minnesota. You do not need this study to apply. You might want it so you go in with the same numbers the regulator published.

Start with the condition page we already keep live: medical cannabis for cancer in Minnesota. Then the qualifying conditions list if you are checking more than one box, and the medical card guide for cost, process, and certification. Medical purchases are taxed differently than adult-use. The medical vs recreational tax explainer is the short version of that math.

Shop the dispensary directory and price tracker once you have a card, not before you have talked to a clinician about dose. If you want the market sitting next to the medicine, the September state of the market is the current official sales snapshot.

Disclosure: the medical card guide includes affiliate links to certification providers. If you sign up through one, MN Cannabis Hub may earn a commission at no extra cost to you. It does not change which numbers we print from OCM.

FAQ

Did Minnesota prove cannabis treats cancer?

No. This is symptom reporting from people already in the medical program. OCM did not measure tumors. It measured nausea, vomiting, pain, appetite, sleep, anxiety, depression, and fatigue on a 0 to 10 scale.

Why do the four-month numbers look better than the "kept it" numbers?

People leave. The report says loss to follow-up is a major limitation. The last column in the table above is the conservative number: improved and still improved four months later, as a share of everyone who started moderate-to-severe.

Is this only Twin Cities patients?

No, but the Twin Cities ZIP regions are 72.9 percent of the cohort (Minneapolis 42.5, St. Paul 30.4). St. Cloud is next at 6.1 percent. Rochester is 3.9 percent.

Do I need a medical card if adult-use is legal?

Not to buy, if you are 21. A card is about potency access, tax treatment, and protections adult-use buyers do not have. Whether that is worth it depends on your condition and how much you buy. We walk through that on the medical card page.

Where do I read the original?

The report PDF is on OCM's site: Impact of Medical Cannabis on Minnesota Cancer Patients. The news release is the September 11 summary. If a number in this article disagrees with a screenshot of the release, trust the PDF table.

Sources

• Minnesota Office of Cannabis Management, Division of Medical Cannabis. Impact of Medical Cannabis on Minnesota Cancer Patients. September 10, 2026. PDF. Tables 2.1, 2.3, 2.4, 2.6, 4.1A; executive summary; side-effects section; product counts.
• Minnesota Office of Cannabis Management. News release, September 11, 2026. Permalink. Used for the Christensen and Zylla quotations and for OCM's program-size claims, which are attributed to OCM rather than treated as Hub findings.