Cannabis and Multiple Sclerosis: What the Evidence Actually Shows

Twenty-five randomized trials. Three thousand seven hundred and sixty-three participants. Four decades of research on one question.

And the same result keeps coming back: people with MS report that cannabis reduces their spasticity, while the scales clinicians use to measure spasticity barely move.

That gap is the entire story, and it is more interesting than either side of the argument usually makes it sound.

This guide covers what the trials found, why patient reports and objective measures diverge, the cognitive cost that rarely gets mentioned, and what is actually approved in the United States.

The short answer

The best evidence supports modest relief of patient-reported spasticity and pain. It does not support any effect on the disease itself, and it comes with a documented cognitive tradeoff.

We sell cannabis seeds. We do not sell medicine, we are not clinicians, and nothing below is advice about managing MS. Anyone with MS considering cannabis should be having that conversation with their neurologist, who will also know about interactions with the drugs they have already prescribed.

What the trials found

The Cochrane review published in 2022 is the most careful summary available. It pooled 25 randomized controlled trials covering 3,763 participants, of whom 2,290 received cannabinoids.

Its central finding on nabiximols, a standardized THC and CBD spray, is that it probably increases the number of people reporting an important reduction in perceived severity of spasticity, with an odds ratio of 2.51, amounting to 216 more people per 1,000 compared with placebo. That is rated moderate-certainty evidence.

The same review states plainly that there is no high-quality evidence in this field.

Read those together and you get a fair summary of forty years of work. Something real is happening for a meaningful minority of people, and the research is not strong enough to say much more.

The gap between feeling and measuring

The American Academy of Neurology's guideline on this is the clearest illustration, and its structure is worth understanding.

It gives its strongest rating, Level A, to oral cannabis extract for reducing patient-reported symptoms of spasticity and pain. In the same document it rates oral cannabis extract as probably ineffective for improving objective spasticity measures. Synthetic THC and nabiximols get the same split. Smoked cannabis is rated as having inadequate data either way.

Every positive recommendation is about what patients say. Every negative one is about what instruments record.

The big British trials show the same pattern in raw form. CAMS, which enrolled 630 people at 33 centers, found no treatment effect on its primary outcome, an objective muscle tone scale, while finding a clear effect on patient-reported spasticity and pain. MUSEC, which used a patient-reported primary endpoint, came out positive, with roughly twice the rate of relief on cannabis extract as on placebo. CUPID, testing whether THC slowed disability progression, failed on every outcome.

A 2024 meta-analysis put numbers on the divergence, finding the pooled effect on patient-reported scores roughly three and a half times larger than the effect on the objective scale.

There are two honest readings of that. Either the scales are measuring something other than what bothers patients, which many clinicians believe, or subjective reports in unblindable trials run high, which is also true. Both can be happening at once.

The one trial that used actual flower

Nearly all of this research used pharmaceutical extracts. One notable exception did not, and it is the most useful trial for anyone reading a cannabis website.

Researchers at the University of California San Diego ran a randomized placebo-controlled crossover study in people with treatment-resistant spasticity, using smoked cannabis cigarettes at around 4 percent THC. On the objective muscle tone scale, the one that usually shows nothing, spasticity fell by an average of 2.74 points more than placebo, and pain scores fell by 5.28 points more.

That is a positive result on the hard endpoint. It is also the trial with the clearest evidence of the cost, which is covered next.

Note the potency. Four percent THC is a fraction of what a modern dispensary sells, and it produced a measurable effect. Nothing in this literature suggests that stronger is better here.

The unexamined day is a wasted opportunity. Reflect on what you did, what you learned, and how you can improve.

John Dewey

The cognitive cost

Most cannabis writing about MS leaves this part out, which is indefensible, because MS itself affects cognition in a large share of patients.

In that same UCSD trial, cognitive processing scores dropped by 8.67 points more on cannabis than on placebo. Benefit and impairment came out of one dataset, in one group of people, at the same time.

The larger finding came from a Toronto group who took 40 people with MS who had used cannabis heavily for years and who were already cognitively impaired, and had half of them stop. By day 28, the withdrawal group performed significantly better on every cognitive index, with brain imaging showing increased activation in the regions associated with the task. Assignment was by odd and even case numbers instead of true randomization, so treat it as strong evidence and not proof.

A 2025 study of 847 people with MS at a single clinic found regular cannabis users performing worse on processing speed and working memory than non-users. That one is cross-sectional, so it shows association and not cause.

Even the National MS Society's own clinician guidance, which is broadly supportive on spasticity and pain, states that chronic whole-plant cannabis adversely affects cognitive performance, while noting that short-term medicinal use does not appear to.

Nothing is approved for MS in the United States

The National Center for Complementary and Integrative Health puts it in one sentence: no marijuana-derived medications are approved by the US Food and Drug Administration to treat MS.

Nabiximols, sold as Sativex, is approved in the United Kingdom, Canada and much of Europe for moderate to severe MS spasticity in patients who have not responded to other treatment. It is not approved in the United States, and the American phase 3 trial that might have changed that missed its primary endpoint in 2022.

The cannabis-derived drugs the FDA has approved treat other things entirely. Epidiolex is for seizures in three specific syndromes. Dronabinol and nabilone are for chemotherapy nausea and AIDS-related weight loss. None of them is indicated for MS.

CBD on its own has never worked here

This is the single most useful thing we can tell a reader, and it cuts against most of what gets marketed.

Across every trial named above, and across the 25 in the Cochrane review, the interventions were THC alone or THC combined with CBD. Not one positive result for MS spasticity in this evidence base comes from CBD by itself.

That is an absence of evidence and not evidence of absence. Nobody has run the trial properly. But anyone being sold CBD oil for MS spasticity should know that the research supporting cannabinoids in MS is research on THC.

People are using it anyway

A survey of the largest US MS registry, with 3,249 respondents, found that 31 percent had used cannabis to treat MS symptoms and 20 percent were currently using it, most commonly for spasticity at 80 percent, pain at 69 percent and sleep at 61 percent.

The reasons given by the 69 percent who had not tried it are just as telling: not enough data on efficacy, not enough on safety, plus legality and cost.

MS is a qualifying condition in most state medical programs, and 41 states, three territories and the District of Columbia allow medical cannabis. Federal law has barely moved. The April 2026 rescheduling covered FDA-approved products and state-licensed medical marijuana only, and broader rescheduling was still pending as of this summer.

What we would actually say

Our contribution here is not medical, so here is the one thing a breeder genuinely knows something about.

MS commonly involves fatigue and reduced hand function, and a cannabis grow is physical work. Topping, training, defoliating and a long flower cycle are all demands on someone who may not have a reliable amount of energy on a given day. That makes the choice of plant a practical question and not a preference.

What matters is a plant that tolerates being left alone. Short, sturdy, forgiving of an imperfect week, and finished in a predictable window. Anything that needs constant intervention is the wrong plant for a household where good days and bad days are not scheduled.

Garlic Cookies is a 90% indica, Girl Scout Cookies crossed with Chemdawg at 30% THC, flowering in 63 to 70 days at 100 to 120 cm indoors, with garlic and diesel over sweet earth, and effects the plant's own profile describes as comprehensive physical relaxation. Ice Cream Cake is gentler at 27%, Gelato #33 crossed with Wedding Cake, 55 to 65 days, creamy vanilla and earthy pine, calming without much weight.

Both are stable, undemanding plants. Neither is a treatment, and the evidence above says the potency number is not what matters anyway.

The short version

Twenty-five randomized trials support modest relief of patient-reported spasticity and pain, rated moderate certainty at best, with no high-quality evidence in the field. Objective spasticity scales mostly do not move, and the one trial using smoked flower at 4 percent THC is the notable exception.

The cognitive cost is documented. People with MS who stopped long-term heavy use improved on every cognitive measure within four weeks.

Nothing cannabis-derived is FDA-approved for MS in the United States, and every positive result in this literature involved THC, alone or with CBD, never CBD by itself. Roughly one in five people with MS in the largest US registry uses cannabis regardless.

Barney's Farm has been developing premium cannabis genetics since the 1980s, with over 40 Cannabis Cup wins. Explore our full seed catalog and find strains bred for every climate and skill level.