Delta-9-tetrahydrocannabinol (THC).

THC is not the same as cannabis

Delta-9-tetrahydrocannabinol, or THC, is one compound in cannabis. A cannabis product may also contain CBD, other cannabinoids, terpenes, contaminants and varying amounts of active ingredients. A purified or synthetic medicine can be made and studied as a defined product; an inhaled or edible plant product may be absorbed differently and have a different onset and duration.

A strain name does not describe the whole treatment. How a product is taken also affects exposure: an edible may be processed differently from an inhaled product, and an extract may contain other active ingredients. Keep the product and route attached to the evidence.12

From receptor effects to outcomes

THC activates cannabinoid receptors, including CB1 receptors in the central nervous system. This helps explain intoxication and effects on attention and memory, and why researchers investigate possible treatments. It does not predict a benefit for every condition. A laboratory receptor result needs controlled evidence in a defined group of people, measuring benefits and harms that matter to them.1

What US approval does and does not establish

The FDA has not approved the cannabis plant itself to treat any disease. Its cannabis overview lists one cannabis-derived prescription product, Epidiolex (purified CBD), and three synthetic cannabis-related products: Marinol and Syndros (dronabinol), and Cesamet (nabilone). A decision about a defined product and use does not transfer to unapproved flower, edibles or dispensary preparations.

US Marinol capsules contain synthetic dronabinol, a form of delta-9-THC. The label specifies use for anorexia associated with weight loss in adults with AIDS, and chemotherapy-related nausea and vomiting when conventional antiemetics have not worked adequately. Syndros is a separate dronabinol product with its own label. Cesamet contains nabilone, a different synthetic cannabinoid chemically similar to THC. The FDA label retrieved describes use for chemotherapy-associated nausea and vomiting after inadequate response to conventional antiemetics. Nabilone is not the same molecule as delta-9-THC. State access arrangements do not themselves establish FDA review of a product’s quality, effectiveness or safety.345

Effects and harms depend on route and context

THC can affect alertness, coordination, memory, perception, heart rate and mood. Inhaled products generally have faster onset than oral products, while oral exposure can be delayed and prolonged; this makes “strength” a poor proxy for blood concentration or impairment. For oral dronabinol, the FDA label describes extensive first-pass metabolism and an active 11-hydroxy-THC metabolite. Absorption and metabolism differ from inhalation, but the label for a capsule does not predict the timing or intensity of every edible product. Alcohol and other sedating substances can add to impairment. Psychiatric, cardiovascular, developmental, pregnancy and medication-related risks may not be represented in small studies. These effects matter because an outcome such as “felt relaxed” may coexist with impaired driving or reduced working memory. Trial reports should state whether participants were instructed about driving, whether adverse effects caused discontinuation, and how long follow-up continued. A delayed or uncommon harm can be missed by a short study. The U.S. Marinol label also warns about psychiatric and cognitive effects, blood-pressure or heart-rate changes, seizures or seizure-like activity, misuse and paradoxical nausea or abdominal pain, and notes that CYP2C9/CYP3A4 inhibitors or inducers can alter dronabinol exposure. These are warnings for a particular prescription product; they do not supply event rates for all cannabis preparations.124

Can you tell when it is safe to drive?

THC can slow reactions and affect coordination and judgement. Feeling capable is not a reliable safety test. In a 2022 US placebo-controlled simulator study of 191 regular cannabis users, participants increasingly rated themselves safe to drive while measured performance was still reduced.

A separate 2026 Canadian crossover study analysed 40 regular users after edible THC or placebo. Some THC exposures worsened lane-position control at the times tested, even when blood concentrations were near or below thresholds used in some driving laws. These studies used particular products and simulator tasks. They cannot establish a universally safe waiting time or blood level for every person and route.

Laws vary, and meeting a legal threshold does not prove driving is safe. Alcohol combined with cannabis can increase impairment. Plan a safe alternative to driving after use. For a separate question, see CBD and driving.678

Evidence is indication-specific

A result for chemotherapy-related nausea does not establish a benefit for chronic pain, sleep, anxiety or cancer itself. Trials may use pharmaceutical products, controlled exposure and monitoring unlike commercial products. Small or short studies may miss rare psychiatric, cardiovascular, dependence or cognitive outcomes. Observational studies can find associations but other factors—such as product choice, tobacco or alcohol use, socioeconomic factors and underlying illness—may also influence results.31

Why evidence packages differ

Medicine trials usually define the product, comparison, eligible participants, outcome and follow-up in advance. Cannabis products used outside trials can vary in composition, potency and use, making it harder to identify which component produced an effect. One observed association cannot separate THC’s effect from why people chose a product or establish that a cannabinoid caused a later diagnosis.31

Keep molecule, formulation and outcome together

“THC works” is too broad. A useful claim names the product, comparison, route, people, outcome, follow-up and uncertainty. It reports harms and treatment stopping as well as benefits. Receptor activity in a laboratory does not show that someone will feel better, function better or live longer.31

Five questions for a THC claim

Is the evidence about dronabinol, nabilone, another purified preparation or a variable cannabis product? Which country and label apply, and does the claim concern access or formal authorisation? Who was studied and for what condition? How were impairment, psychiatric effects, interactions, dependence, treatment stopping and delayed harms assessed? Is the source a controlled trial, observational study, laboratory experiment or anecdote?31

Development & regulatory context

These selected records are not a complete global legal-status list. Approval depends on the product, use, population and country; check each record’s date.