What the evidence actually says about cannabis

Cannabis contains over 100 compounds called cannabinoids, the most studied being THC (which produces the high) and CBD (which does not). Research into health effects is still developing—most human studies are small, many are observational rather than controlled trials, and the evidence for most claimed benefits remains preliminary. What follows is what peer-reviewed research has found so far, separated clearly from what remains unproven.

The strongest evidence exists for cannabis in treating specific types of seizures and for reducing nausea in people undergoing chemotherapy. Moderate evidence supports its use for chronic pain and multiple sclerosis symptoms. For most other claimed benefits—anxiety, sleep, inflammation, appetite—the human research is limited, contradictory, or absent. Animal studies suggest mechanisms that might matter, but animal findings often do not translate to humans.

Key Takeaways

  • Cannabis shows the strongest evidence for treating certain seizure disorders and chemotherapy-related nausea, based on human clinical trials.
  • Chronic pain relief has moderate support from human studies, though the effect size varies widely between individuals.
  • Most other claimed benefits—anxiety, sleep, inflammation—rest on animal studies, small human trials, or observational data, not large controlled trials.
  • THC and CBD have different effects; CBD does not produce intoxication but also has less research behind it than THC for most conditions.
  • Cannabis can interact with medications and may worsen anxiety or psychosis in some people, particularly those with a family history of mental illness.

Seizure disorders and FDA-approved cannabis medicine

The clearest medical use is Epidiolex, a CBD-based medication approved by the FDA in 2018 for two rare, severe seizure disorders: Lennox-Gastaut syndrome and Dravet syndrome. Controlled trials showed CBD reduced seizure frequency by roughly 40 percent in patients who had not responded to standard anti-seizure drugs. This is the only cannabis-derived medicine with FDA approval and the strongest human evidence behind any cannabis claim.

For other seizure types, the evidence is weaker. Some observational studies suggest cannabis may help, but controlled trials are sparse. If you have a seizure disorder, the conversation belongs with a neurologist who knows your specific diagnosis and current medications, not with general cannabis information.

Chronic pain: moderate evidence with high individual variation

Multiple systematic reviews of human trials find that cannabis reduces chronic pain in some people, but the effect is modest and inconsistent. A 2021 review in JAMA found pain relief in about half of trial participants, with average reductions of 30 percent or less. Importantly, placebo effects in pain trials are large—some of the benefit may not be specific to cannabis itself.

The research does not show which types of pain respond best or which patients will benefit. Neuropathic pain (nerve damage) and cancer pain have somewhat more evidence than other types. Most studies used smoked cannabis or oral THC rather than CBD alone, and most lasted weeks to months, so long-term effects remain unclear. People taking opioids should not switch to cannabis without medical supervision, as the interaction is not well understood.

Nausea from chemotherapy and other causes

Cannabis, particularly THC, has solid evidence for reducing nausea and vomiting in people undergoing chemotherapy. Multiple controlled trials show benefit, and the FDA approved dronabinol (synthetic THC) and nabilone (a THC analog) for this specific use decades ago. For other causes of nausea—migraines, gastroenteritis, general post-operative nausea—the evidence is much thinner and mostly anecdotal.

The anti-nausea effect appears to work through cannabinoid receptors in the brain and gut, but the mechanism is not fully mapped. Effectiveness varies by individual and by dose, and some people experience increased nausea or anxiety at higher doses.

Anxiety, sleep, and inflammation: what animal studies suggest but humans have not confirmed

CBD is widely marketed for anxiety and sleep, and animal studies show it can reduce anxiety-like behavior and affect sleep architecture in rodents. However, human trials are small and mixed. A 2019 review found some evidence that CBD may help anxiety, but most studies involved people with specific conditions (social anxiety, PTSD) rather than general anxiety, and sample sizes were small. For sleep, the human evidence is even thinner—mostly case reports and small open-label studies, not controlled trials.

Inflammation is another area where cannabinoids show promise in cell and animal studies, but human evidence is scarce. A few small trials suggest CBD or cannabis may reduce inflammatory markers in people with certain conditions, but these are not large enough to guide treatment decisions. The gap between "works in a petri dish" and "works in a person" is real and often large.

Risks and interactions you should know about

Cannabis is not risk-free. THC can impair memory, attention, and motor control, particularly in regular users and in people whose brains are still developing (roughly before age 25). Heavy use in adolescence is linked to lower IQ and cognitive changes in some studies, though causation is debated. Cannabis can trigger or worsen anxiety, paranoia, and psychosis, especially in people with a personal or family history of schizophrenia or bipolar disorder.

Cannabis interacts with many medications, including blood thinners, seizure drugs, and medications metabolized by the liver. If you take prescription drugs, a pharmacist or doctor should review cannabis use before you start. Smoking cannabis carries the same respiratory risks as smoking anything else. Edibles pose a risk of accidental overdose, particularly for new users, because effects take 1 to 2 hours to appear and people often consume more while waiting.

The difference between THC-dominant and CBD-dominant products

THC is the compound that produces intoxication and has the most research behind it for pain and nausea. CBD does not cause a high and is legal in most U.S. states, but it has far less human research. Many CBD products sold over the counter are not well regulated—third-party testing shows that some contain little CBD, some contain unlabeled THC, and some contain contaminants.

If you are considering cannabis for a medical reason, the difference matters. A product labeled "CBD" may not contain what the label says. Products with both THC and CBD may work differently than either alone, but this interaction is not well studied in humans. Prescription cannabis medicines like Epidiolex and dronabinol are standardized and tested, which over-the-counter products are not.

What questions remain open

Researchers still do not know the optimal dose for most conditions, the best route of administration (smoking, vaping, edibles, oils), or how long effects last with regular use. Long-term safety data in humans is sparse—most trials last weeks to months. We do not know whether cannabis prevents or treats any disease, only whether it may reduce symptoms in some people for some conditions.

The legal status of cannabis varies by state and country, which has slowed research. Many studies are small, funded by cannabis companies or advocacy groups, or conducted in places where rigorous blinding and placebo controls are difficult. Larger, longer, independently funded trials would clarify which claims hold up and which do not.

Frequently Asked Questions

Is CBD safer than THC?

CBD does not cause intoxication and appears to have a good safety profile in short-term studies, but long-term human safety data is limited. It can interact with medications and may cause drowsiness or liver changes at high doses. "Safer" depends on the individual and the condition—neither is risk-free, and neither has been studied as thoroughly as common prescription drugs.

Can cannabis treat cancer?

No human evidence shows cannabis cures or treats cancer itself. Some research suggests cannabinoids may slow cancer cell growth in lab dishes and animals, but this has not translated to human benefit. Cannabis may help manage pain and nausea from cancer treatment, but it should not replace standard cancer therapy.

Will cannabis show up on a drug test?

Standard drug tests detect THC metabolites, not CBD. If you use any product containing THC—including hemp products that claim to be "THC-free" but may contain trace amounts—you could test positive. CBD alone will not trigger a positive result on most workplace or legal drug tests.

Is cannabis addictive?

Regular THC use can lead to psychological dependence and withdrawal symptoms (irritability, sleep problems, anxiety) when stopping, though physical dependence is less severe than with alcohol or opioids. About 9 percent of cannabis users develop cannabis use disorder; the rate is higher in people who start young or use daily. CBD has not been shown to be addictive.

Can I use cannabis instead of my anxiety medication?

Do not stop or replace prescription anxiety medication with cannabis without talking to your doctor. Cannabis may worsen anxiety in some people, and stopping psychiatric medication abruptly can be dangerous. If you want to explore cannabis as part of your treatment, your prescriber needs to know so they can monitor you and adjust other medications if needed.