Does Cannabis Cause Cancer? What the Evidence Shows

Two things are true at the same time, and that is why the answers to this question conflict so sharply. Cannabis smoke contains known carcinogens, and California has listed marijuana smoke as a cause of cancer since 2009. Yet the large evidence reviews have not established that smoking cannabis causes lung cancer or head and neck cancer. The 2017 National Academies report graded the lung evidence as moderate evidence of no association. Understanding how both can hold at once is the whole answer.

Does Cannabis Cause Cancer?

The evidence does not establish that it does, and it does not clear it either.

That sounds evasive, so here is the precise version. Cannabis smoke contains chemicals that are known to cause cancer. When researchers look at populations of cannabis smokers and count their cancers, they have generally not found the excess those chemicals would predict. One association has survived pooled analysis, testicular germ cell tumors after long-term use, and the researchers who found it graded their own evidence as low strength.

The gap between those two facts is not a contradiction. It is the difference between a hazard, meaning an agent capable of causing cancer, and a risk, meaning a measurable excess of cancer in people actually exposed. The rest of this article works through where the evidence sits for each cancer type and why the gap exists.

Does Cannabis Smoke Contain Carcinogens?

Yes, many of the same ones found in tobacco smoke, and several at higher concentrations.

The most direct comparison comes from a 2008 study in Chemical Research in Toxicology that prepared cannabis and tobacco cigarettes the same way and burned them under two matched machine smoking conditions. Its findings:

The authors described the results as qualitatively similar with quantitative differences, and concluded that confirming known carcinogens in both mainstream and sidestream cannabis smoke was important public health information.

California acted on that class of evidence. The state's Proposition 65 list, published by the Office of Environmental Health Hazard Assessment, carries marijuana smoke under the toxicity type cancer with a listing date of June 19, 2009. A separate entry, cannabis (marijuana) smoke, was listed for developmental toxicity on January 3, 2020. That is a different endpoint from cancer, and it is noted here only so the two entries are not confused.

A Proposition 65 listing is a hazard identification. It says the agent is capable of causing cancer. It does not quantify how much cancer occurs at real-world exposure levels, and that is precisely the question the epidemiology below tries to answer.

Has IARC Classified Cannabis Smoking as a Carcinogen?

No. As of August 2026 no classification exists, and one is scheduled.

The International Agency for Research on Cancer is the World Health Organization body that assigns the familiar Group 1, 2A, and 2B carcinogen classifications. Cannabis smoking is the sole agent scheduled for IARC Monographs Volume 143, with the working group convening in Lyon, France from 3 to 10 November 2026.

This is worth stating plainly because a great deal of online content asserts an IARC classification for cannabis that has never been issued. If you encounter a page citing a group number, check it against the IARC list directly. The evaluation is pending, not complete.

Does Smoking Cannabis Cause Lung Cancer?

The research has not shown that it does, and this is the most studied question of the set.

The 2017 National Academies of Sciences, Engineering, and Medicine report reviewed the literature and issued graded conclusions. On lung cancer, Conclusion 5-1 states there is moderate evidence of no statistical association between cannabis smoking and the incidence of lung cancer.

The National Cancer Institute reaches the same reading, stating that observational studies failed to demonstrate a statistically significant association between cannabis inhalation and lung cancer after adjusting for tobacco use.

The 2019 meta-analysis in JAMA Network Open went further and declined to pool the lung cancer studies at all. Its authors described the findings as mixed and confounded by few marijuana-only smokers, poor exposure assessment, and inadequate adjustment.

One qualifier belongs alongside that. The National Cancer Institute notes a pooled analysis in which tobacco smokers who also inhaled cannabis showed a significantly increased risk of lung cancer. Co-use is common, and the reassurance about cannabis alone does not extend to people who also smoke tobacco.

Read carefully, "moderate evidence of no association" is a statement about what studies have detected. It is not a finding that smoke is harmless. The National Academies applied a separate set of conclusions to respiratory harm, and there it found substantial evidence linking long-term cannabis smoking to worse respiratory symptoms and more frequent chronic bronchitis.

What About Head and Neck Cancer?

Here the evidence has moved since 2017, and the two pictures are worth seeing side by side.

The National Academies Conclusion 5-2 found moderate evidence of no statistical association between cannabis use and the incidence of head and neck cancers. The 2019 JAMA Network Open meta-analysis agreed, finding no association for ever use:

• Head and neck squamous cell carcinoma, odds ratio 1.26 (95% CI 0.88 to 1.80)
• Oral cancer, odds ratio 1.22 (95% CI 0.95 to 1.56)

Neither reached statistical significance. Then in 2024 a much larger study reported something different. A cohort study in JAMA Otolaryngology drew on 20 years of records from 64 health care organizations and used propensity score matching for demographics, alcohol-related disorders, and tobacco use. After matching 115,865 people per group:

Those are large numbers, and the exposure they describe is the key to reading them. The study measured cannabis-related disorder, a clinical diagnosis of problematic use recorded in a medical chart, not cannabis use in general. People who carry that diagnosis are not a random sample of people who use cannabis, and a database of clinic visits captures people already engaged with the health system.

The authors said as much. Their stated conclusion calls for future research to examine the mechanism and analyze dose response with strong controls, given the limitations of the database.

So the honest summary is narrow. A large recent study found a strong association in a heavily-using clinical population. It has not yet been reconciled with the earlier pooled analyses of ordinary use. That is an open question, not a settled reversal.

Is There a Link to Testicular Cancer?

This is the one association that survived pooling, and it applies to long-term use rather than any use.

National Academies Conclusion 5-3 found limited evidence of a statistical association between current, frequent, or chronic cannabis smoking and non-seminoma-type testicular germ cell tumors. Note the direction. On lung and head and neck cancer the report found evidence of no association. Here it found evidence of an association, at its weaker grade.

The 2019 meta-analysis quantified it across three case-control studies:

• Ever use and testicular germ cell tumors, odds ratio 1.11 (95% CI 0.81 to 1.53). No association.
• More than ten years of use, odds ratio 1.36 (95% CI 1.03 to 1.81), with no heterogeneity between studies.
• More than ten years of use and the nonseminoma subtype, odds ratio 1.85 (95% CI 1.10 to 3.11).

The authors graded this low strength evidence, and they framed their overall conclusion around it. Low-strength evidence suggests that smoking marijuana is associated with developing testicular germ cell tumors. Its association with other cancers, and the consequences of heavier use, remain unclear.

The National Cancer Institute adds a cohort finding pointing the same way. In a 42-year Swedish follow-up, heavy cannabis use, defined as more than 50 times in a lifetime, was associated with a 2.5-fold increased risk.

Testicular germ cell tumors are uncommon, and a relative increase on a small baseline is a small absolute change. The pattern is consistent enough to take seriously and thin enough that nobody should be quoted a personal risk number from it.

What About Other Cancers?

For nearly everything else, the National Academies found the evidence insufficient to support or refute an association.

Conclusion 5-4 covers esophageal cancer. Conclusion 5-5 covers prostate cancer, cervical cancer, malignant gliomas, non-Hodgkin lymphoma, penile cancer, anal cancer, Kaposi's sarcoma, and bladder cancer. Conclusion 5-6 covers parental cannabis use and a subsequent risk in offspring of acute myeloid leukemia, acute non-lymphoblastic leukemia, acute lymphoblastic leukemia, rhabdomyosarcoma, astrocytoma, or neuroblastoma.

Insufficient evidence means the studies are too few, too small, or too flawed to answer the question. It does not mean no risk, and it does not mean risk. It means nobody knows.

One finding worth flagging inside that category: the National Cancer Institute notes that among tobacco non-smokers, ever having used cannabis was associated with an increased risk of prostate cancer. That sits inside Conclusion 5-5's insufficient grade rather than overturning it.

Why Is the Evidence So Weak After Fifty Years?

Because the study designs available have specific, well-documented problems that a larger sample size does not fix.

The 2019 meta-analysis screened studies published from 1973 onward and included 25. Of those, 19 were case-control, 5 were cohort, and 1 was cross-sectional. Only 2 of the 25 were at low risk of bias. The recurring problems the authors identified:

• Tobacco confounding. Most cannabis smokers in these studies also smoked tobacco. Isolating a cannabis effect from a far stronger and far better documented carcinogen is difficult, and cohorts of cannabis-only smokers are small.
• Exposure measurement. The meta-analysis defined ever use as at least one joint-year, meaning one joint per day for a year. Real exposure histories are self-reported, rarely quantified, and often recalled decades later.
• Few heavy, long-term users. The authors specifically cited a small number of participants with high levels of use.
• Young populations and latency. Study populations were young, and few studies measured longitudinal exposure. Solid tumors typically take decades to appear, so a cohort followed from its twenties into its forties may simply be too early.

Dose is the likely explanation for the lung finding specifically. A person who smokes a pack a day for thirty years accumulates an exposure that a typical cannabis smoker does not approach. Carcinogens present at comparable or higher concentrations per cigarette can still produce no detectable population signal if the total lifetime exposure is an order of magnitude lower.

Does the Route of Administration Matter?

Mechanistically it should matter a great deal, and no study has measured cancer outcomes by route.

Every carcinogen described in this article comes from combustion. Ammonia, hydrogen cyanide, aromatic amines, and polycyclic aromatic hydrocarbons are products of burning plant material. An edible, a tincture, or a capsule produces none of them.

That is a strong mechanistic argument and it is not an outcome finding. Nobody has followed a cohort of people who used only non-combusted cannabis for the decades required to count their cancers. The National Academies conclusions on lung and esophageal cancer specify cannabis smoking, which reflects what the underlying studies measured.

So the accurate statement is narrow: the exposure that carries the identified carcinogens is absent from non-combusted products, and the outcome research to confirm what that means does not exist.

Is Cancer a Qualifying Condition in Florida, Georgia, and Texas?

Yes. Cancer is a listed qualifying condition in all three state programs.

• Florida. Cancer is enumerated in Fla. Stat. 381.986.
• Georgia. The Georgia Department of Public Health Low THC Oil Registry conditions under O.C.G.A. 31-2A-18 list any cancer except skin cancer, unless the skin cancer is metastatic.
• Texas. Cancer is listed in Occupations Code 169.003 as amended by HB 46 of 2025.

That is a different question from the one this article answers. This article is about whether cannabis use causes cancer. Whether cannabis helps people who already have a cancer diagnosis, what the evidence supports for treatment-related symptoms, and how eligibility works in each state are covered on the cancer condition page. Nothing here should be read as guidance on using cannabis during cancer treatment, which is a conversation for an oncology team.

What This Means for Florida, Georgia, and Texas Patients

Combustion is the part with identified carcinogens. If cancer risk is your reason for asking, that is the specific exposure the research is about, and it is worth raising route of administration at the appointment.

Tell the physician about tobacco use. The clearest cancer signal involving cannabis in the lung literature appears in people who also smoke tobacco. That history changes the picture more than the cannabis question does.

Do not read "no association found" as a clean bill of health. The same National Academies report that found moderate evidence of no association with lung cancer found substantial evidence linking long-term cannabis smoking to worse respiratory symptoms and more frequent chronic bronchitis. Cancer is one outcome, not the only one.

Long-term heavy use is where the signals cluster. Both the testicular finding and the 2024 head and neck study describe long-duration or clinically problematic use, not occasional use. Duration and intensity are the variables that matter in the data.

Bring your family history. None of the research above is personalized. A physician weighing your history can say more about your situation than a population odds ratio can.

Miracle Leaf® physicians evaluate patients for the Florida, Georgia, and Texas state programs.

What If You Live Outside Florida, Georgia, and Texas?

The research in this article is not state specific. The certification pathway is.

If you are in a state where medical cannabis is legal but Miracle Leaf has no clinic, the telehealth program covers evaluations in states outside the Florida, Georgia, and Texas footprint. The same points apply: route of administration, tobacco history, and duration of use all belong in the conversation.

Sources on Cannabis and Cancer Risk

• The Health Effects of Cannabis and Cannabinoids, Chapter 5: Cancer. National Academies of Sciences, Engineering, and Medicine. Washington DC: National Academies Press, 2017. doi:10.17226/24625. Source of graded Conclusions 5-1 through 5-6.
• The Health Effects of Cannabis and Cannabinoids, Chapter 7: Respiratory Disease. National Academies of Sciences, Engineering, and Medicine, 2017. Source of Conclusion 7-3(a), substantial evidence of a statistical association between long-term cannabis smoking and worse respiratory symptoms and more frequent chronic bronchitis episodes.
• Association Between Marijuana Use and Risk of Cancer: A Systematic Review and Meta-analysis. Ghasemiesfe M, Barrow B, Leonard S, Keyhani S, Korenstein D. JAMA Netw Open. 2019;2(11):e1916318. doi:10.1001/jamanetworkopen.2019.16318. PMCID PMC6902836. Source of the head and neck, oral, and testicular germ cell tumor pooled estimates, the evidence-strength grades, and the risk-of-bias counts.
• Cannabis Use and Head and Neck Cancer. Gallagher TJ, Chung RS, Lin ME, Kim I, Kokot NC. JAMA Otolaryngol Head Neck Surg. 2024;150(12):1068 to 1075. doi:10.1001/jamaoto.2024.2419. PMCID PMC11310842. Source of the cannabis-related disorder cohort sizes and site-specific relative risks.
• A comparison of mainstream and sidestream marijuana and tobacco cigarette smoke produced under two machine smoking conditions. Moir D, Rickert WS, Levasseur G, et al. Chem Res Toxicol. 2008;21(2):494 to 502. doi:10.1021/tx700275p. Source of the smoke constituent comparison.
• Cannabis and Cannabinoids (PDQ), Health Professional Version. NIH National Cancer Institute. Source of the lung cancer adjustment statement, the co-use pooled analysis, the 42-year Swedish testicular follow-up, and the prostate cancer observation.
• Proposition 65 List of Carcinogens or Reproductive Toxicants. California Office of Environmental Health Hazard Assessment, list dated July 31, 2026. Marijuana smoke, toxicity type cancer, listed June 19, 2009. Cannabis (marijuana) smoke, toxicity type developmental, listed January 3, 2020.
• IARC Monographs Volume 143: Cannabis Smoking, Call for Experts. International Agency for Research on Cancer. Working group meeting Lyon, France, 3 to 10 November 2026. Confirms no IARC evaluation of cannabis smoking has been completed.
• Fla. Stat. 381.986, Georgia Low THC Oil Registry conditions under O.C.G.A. 31-2A-18, and Texas HB 46 of 2025 amending Occupations Code 169.003. State qualifying condition lists.

Related Health and Eligibility Resources

• The cancer condition page covers eligibility in all three states and what the evidence shows for patients who already have a cancer diagnosis.
• The health benefits of medical cannabis post covers the conditions where the evidence base is strongest.
• The does weed raise blood pressure post covers the cardiovascular research, which follows a similar pattern of association without established causation.
• The is weed a depressant, stimulant, or hallucinogen post covers the pharmacology underneath these effects.
• The qualifying conditions page covers eligibility across all three state programs.
• The Florida marijuana laws, Georgia marijuana laws, and Texas marijuana laws pages cover each program and its condition list.

Talk to a Physician About Your Risk History

Family history, tobacco use, duration of cannabis use, and route of administration all change what a physician can tell you. Bring them to the appointment.

Call (833) LEGAL-MJ or contact us online to book. Evaluation pricing is on the pricing page, and the clinicians who review this content are listed on the editorial team page.

Disclaimer

This post is informational and is not medical advice. It addresses whether cannabis use causes cancer and does not address cannabis as a cancer treatment or as supportive care during treatment. The figures cited come from published research populations and do not predict outcomes for any individual. The pooled estimates are drawn largely from case-control studies, only 2 of 25 of which were rated at low risk of bias, with acknowledged tobacco confounding and poorly quantified exposure. The 2024 head and neck study measured diagnosed cannabis-related disorder rather than cannabis use generally, and its authors state that dose response work with stronger controls is still needed. No International Agency for Research on Cancer evaluation of cannabis smoking has been completed as of August 2026. Consult a licensed physician about cannabis use in the context of your own history, and consult your oncology team about any cannabis use during cancer treatment.