According to a YouGov survey of 1,148 US adults, 64 percent of Americans consider regular alcohol use more dangerous than cannabis, while 63 percent said the same about tobacco.
Only 14 percent said marijuana was more problematic than alcohol.
A separate 2025 MRI-Simmons National Cannabis Study found that 61 percent of Americans agreed that marijuana is a healthier choice than alcohol, rising to 87 percent among people who had used cannabis in the past six months.
A Gallup poll released in August 2024 reinforced the trend, finding that Americans ranked marijuana lowest in perceived harmfulness out of eight commonly used substances, placing it well below alcohol, cigarettes, vapes, and chewing tobacco.
These are not fringe opinions.
They represent a fundamental and accelerating shift in how the American public understands two of the most widely used substances in the country.
The question worth asking, carefully, is how much of that shift is supported by the science and how much of it is running ahead of what researchers actually know.
How Public Perception Has Shifted and Why
The transformation in how Americans view cannabis has been one of the most dramatic reversals in public opinion in recent history.
A generation ago, cannabis was widely understood, largely thanks to programs like D.A.R.E., as a dangerous gateway drug with no legitimate health applications.
Today, recreational marijuana is legal in 24 states and medicinal use is permitted in 39, and more adults now smoke cannabis than smoke cigarettes, a shift documented in federal survey data for the first time in 2024.
As NPR reported, nearly 18 million Americans aged 12 and older reported daily or near-daily cannabis use in 2022, compared to approximately 15 million who said the same about alcohol.
That inversion, daily cannabis use overtaking daily alcohol use, would have been unimaginable a decade ago.
Dr. Godfrey Pearlson, professor of psychiatry and neuroscience at Yale University, told NPR that the shift in perception is directly linked to the shift in use.
“Perception of safety is definitely linked epidemiologically to willingness to consume any drug,” he said.
As legalization has expanded, contact with cannabis has normalised.
As normalisation has grown, fear has retreated.
And as fear has retreated, use has risen.
Whether that rising use is accompanied by an accurate understanding of the risks involved is a different and more complicated question.
Where the Science Actually Agrees With Public Perception
On several of the most important measures of harm, the public perception that cannabis is less dangerous than alcohol is supported by the research.
Alcohol is a known carcinogen.
The National Cancer Institute lists alcohol as a Group 1 carcinogen, directly linked to cancers of the mouth, throat, esophagus, liver, colon, rectum, and breast.
Cannabis does not carry the same established cancer association.
The US National Academy of Sciences Institute of Medicine found no conclusive evidence that marijuana causes cancer in humans, including cancers usually associated with tobacco use.
On liver toxicity, the comparison is equally stark.
Chronic alcohol use is one of the leading causes of liver disease worldwide, responsible for alcoholic hepatitis, fibrosis, and cirrhosis.
Cannabis has no established equivalent effect on liver function in people who do not already have liver disease.
On violence and aggression, the research consistently points in one direction.
As research published in the journal Addictive Behaviors found, “alcohol is clearly the drug with the most evidence to support a direct intoxication-violence relationship,” while cannabis “reduces the likelihood of violence during intoxication.”
Alcohol is also dramatically more toxic in overdose.
Alcohol poisoning kills thousands of Americans annually.
There are no documented cases of fatal cannabis overdose in humans.
On impaired driving, the comparison is more nuanced but still generally favours cannabis.
Dr. Pearlson told NPR that cannabis-impaired driving carries roughly a 10 percent increase in crash risk, compared to 10 to 20 times the baseline risk at the legal alcohol limit.
Cannabis impairs driving.
That matters and should not be minimised.
But the magnitude of that impairment is substantially lower than alcohol’s.
What Most People Get Wrong: Cannabis Is Not Risk-Free
Here is where the popular narrative runs significantly ahead of the evidence.
The conclusion that cannabis is safer than alcohol does not mean cannabis is safe.
And for specific populations, the risks of regular cannabis use are serious enough to warrant considerably more caution than public perception currently reflects.
The cardiovascular picture, in particular, has become more concerning as research has accumulated.
A study published in the Journal of the American Heart Association found that daily marijuana use raises the risk of stroke by 42 percent and heart attack by 25 percent, even among people with no prior history of heart disease who had never smoked tobacco.
That is a significant cardiovascular signal that is largely absent from the popular “cannabis is safe” narrative.
Cannabis has also been linked to cardiac arrhythmias, myocarditis, and coronary artery spasm.
As CNN reported in its analysis of the cannabis-versus-alcohol comparison, “cannabis smoke contains toxins, carcinogens, and particulate matter that have been linked to cancer, lung damage, and cardiovascular disease.”
Cannabis smoke may actually be more harmful to the respiratory system than tobacco in one specific respect: users typically hold the smoke in their lungs longer, increasing exposure to combustion byproducts.
A 2023 study found that 75 percent of marijuana smokers showed evidence of emphysema, compared to 67 percent of tobacco smokers and 5 percent of non-smokers.
Those numbers deserve more public attention than they currently receive.
The mental health risks of cannabis, particularly for young people, are also considerably more serious than casual cultural attitudes suggest.
According to the US Centers for Disease Control and Prevention, cannabis use before the age of 25 can permanently damage the developing brain, affecting the capacity to learn, remember, solve problems, and sustain attention.
High-potency cannabis has been linked to psychosis, with research published in The Lancet Psychiatry estimating that high-potency cannabis use may account for 10 percent of new psychosis cases.
Up to 50 percent of those who develop cannabis-related psychosis may later be diagnosed with schizophrenia.
And a brain imaging study published in the Journal of Alzheimer’s Disease found that cannabis abuse was second only to schizophrenia in accelerating brain aging, associated with approximately 2.8 years of premature brain aging, compared to 0.6 years for alcohol abuse.
That is a finding that directly contradicts the popular perception that cannabis is easier on the brain than alcohol.
How the Study Applies to Real Life
The most practically useful takeaway from the accumulated research is not a simple declaration about which substance “wins.”
It is a more nuanced understanding of which harms are concentrated where, and in whom.
For middle-aged adults who drink heavily and are considering reducing or replacing alcohol consumption, the evidence broadly supports the idea that moderate cannabis use carries fewer acute and long-term risks than heavy alcohol use.
Alcohol’s cancer risk, liver toxicity, contribution to violence, and overdose potential make it a genuinely more harmful substance for the average heavy drinker.
That comparison shifts significantly for specific groups.
For teenagers and young adults under 25, cannabis poses serious developmental risks that alcohol, while also harmful to developing brains, does not replicate in the same way.
For people with a personal or family history of psychosis or schizophrenia, cannabis carries specific risks that make it considerably more dangerous than the general population data suggests.
For anyone with existing cardiovascular disease, the 2024 heart attack and stroke data represents a genuine warning signal that has not yet filtered into mainstream awareness.
For pregnant women, the CDC warns that there is no known safe level of cannabis use during pregnancy, with evidence of risk to fetal brain development.
The method of consumption also matters enormously in ways that public perception largely ignores.
Edibles and tinctures eliminate the respiratory risks of combustion entirely.
Vaping reduces some risks but introduces others.
Smoking, particularly daily smoking of high-THC products, carries the highest risk profile and is also the most common consumption method.
The THC Potency Problem Nobody Is Talking About
One dimension of the cannabis safety debate that receives almost no attention in popular culture involves the dramatic increase in THC potency over the past two decades.
Much of the research suggesting cannabis is relatively safe was conducted when the average THC content of cannabis products was significantly lower than it is today.
As Healthline’s medically reviewed June 2025 analysis noted, cannabis may appear safer than alcohol in some comparisons simply because researchers are not yet fully aware of certain risks associated with today’s high-potency products.
The cannabis available at legal dispensaries today often contains THC concentrations of 20 to 30 percent or higher, compared to the 3 to 4 percent typical of cannabis studied in research from the 1990s and early 2000s.
Concentrates and vape cartridges can reach 80 to 90 percent THC.
The safety profile of those products has simply not been studied over the long term in the way alcohol has.
As Professor Kent Hutchison of the University of Colorado told UCHealth, “absence of evidence is not evidence of absence.”
The fact that researchers have not yet definitively established certain harms does not mean those harms do not exist.
It means the research has not caught up with the product.
A Sober Assessment of a Complex Question
The public perception that cannabis is less dangerous than alcohol is, in several important respects, grounded in real and well-documented evidence.
Alcohol kills more people, causes more organ damage, contributes more to violence, and carries a more clearly established cancer risk than cannabis does.
Those are facts, not advocacy.
But the leap from “less dangerous than alcohol” to “safe” is one the evidence does not support.
And the specific populations most likely to be harmed by cannabis, teenagers, people with psychosis risk, heavy daily users, pregnant women, and people with cardiovascular disease, are often the same populations least likely to be reached by nuanced public health messaging.
As legalization continues to expand and cannabis becomes more deeply embedded in mainstream wellness culture, the quality of public understanding about its actual risk profile matters more, not less.
Replacing one oversimplification with another, from “marijuana is a dangerous gateway drug” to “marijuana is safe and healthy,” does not serve the people most at risk.
What they deserve is the same thing everyone deserves: the full picture, delivered honestly, without the distortion of either stigma or marketing.
Cannabis Use Disorder: The Risk Nobody Wants to Talk About
One of the most consistent blind spots in the public conversation about cannabis safety involves addiction.
Because cannabis does not produce the physical withdrawal symptoms most associated with alcohol or opioids, many users assume it is not addictive in any meaningful sense.
The clinical evidence says otherwise.
According to the National Institute on Drug Abuse, approximately 30 percent of people who use cannabis regularly develop cannabis use disorder, a clinical condition characterised by continued use despite significant negative consequences to health, relationships, work, and daily functioning.
That is roughly 3 in 10 regular users.
By comparison, the dependency rate for alcohol is approximately 15 percent of users, meaning that on this specific measure, cannabis may actually carry a higher risk of dependency than alcohol does.
The 2024 SAMHSA National Survey on Drug Use and Health found that rates of cannabis use disorder were 3.7 times higher in 2024 than in 2015, a dramatic increase that tracks closely with the rise of high-potency products and the normalisation of daily use.
According to National Drug and Alcohol Research data, over 16 million Americans currently meet the diagnostic criteria for cannabis use disorder.
That figure is rarely part of the popular conversation about cannabis safety.
Research published in Drug and Alcohol Dependence found that cannabis use disorder frequently co-occurs with other mental health conditions and that despite its high prevalence, only a small proportion of those affected seek or receive treatment, partly because the cultural framing of cannabis as a low-risk substance makes it harder for users to recognise when their use has become a problem.
The frequency data adds further context.
A landmark 2024 Carnegie Mellon study published in the journal Addiction found that the median cannabis user in 2022 reported using on 15 to 16 days per month, compared to the median drinker who reported alcohol use on just 4 to 5 days per month.
In other words, while far more people drink than use cannabis, those who use cannabis tend to use it much more frequently.
Professor Jonathan Caulkins of Carnegie Mellon, who conducted the study, described the finding as striking: “High-frequency cannabis use is now more commonly reported than high-frequency drinking.”
That shift in use patterns has significant implications for how we assess risk.
Many of the studies suggesting cannabis is relatively safe were based on populations using it occasionally.
The health consequences of daily or near-daily use, particularly of high-potency modern products, are considerably less well understood.
Can Cannabis Replace Alcohol? A Cautious Look at an Emerging Trend
One of the more intriguing developments in the cannabis and alcohol conversation involves a growing body of people who are consciously using cannabis as a substitute for alcohol, either to reduce drinking or to eliminate it entirely.
Anecdotally, the trend is well documented.
Alcohol consumption among younger Americans has been declining for years, while cannabis use has surged.
A February 2026 study from the University at Buffalo found that cannabis users who consumed cannabis-infused beverages reported cutting back on alcohol consumption as a result, suggesting that for some people, cannabis beverages may serve as a harm reduction tool by replacing a substance with more clearly established long-term health risks.
For adults who drink heavily and are seeking a lower-risk alternative, that trade-off may be genuinely beneficial on balance.
Alcohol kills nearly 178,000 Americans every year, and for chronic heavy drinkers, the organ damage, cognitive decline, and mortality risk of continued alcohol use is severe and well-established.
Substituting moderate cannabis use for heavy drinking, particularly through non-combustible methods like edibles or tinctures, may represent a net reduction in harm for that specific population.
But the substitution argument has important limits.
For people who do not drink heavily, replacing occasional social drinking with daily cannabis use would almost certainly represent an increase in overall risk, not a reduction.
And the cardiovascular, respiratory, and mental health risks associated with regular cannabis use mean that “safer than heavy alcohol use” is a very specific and narrow benchmark, not a general endorsement of cannabis as a health-positive choice.
What Young People Are Getting Wrong
Perhaps the most consequential gap between public perception and scientific reality involves teenagers and young adults, the demographic whose cannabis use has grown most dramatically alongside falling perceptions of risk.
According to Monitoring the Future survey data, the share of high school seniors who view regular marijuana use as dangerous has fallen from 58 percent in 2000 to just 36 percent in 2024.
In 2025, daily marijuana use exceeded daily cigarette use among 8th graders, 10th graders, and 12th graders, with 5.6 percent of 12th graders reporting daily cannabis use compared to just 0.8 percent reporting daily cigarette use.
This is a generation that has internalised the message that cannabis is safer than tobacco.
What many of them have not internalised is that cannabis, for a developing brain, carries risks that do not apply in the same way to adults.
The CDC is unambiguous on this point: cannabis use before the age of 18 is associated with addiction, psychosis, depression, anxiety, and impaired memory, attention, and learning.
Those risks are not hypothetical or theoretical.
They are documented across multiple longitudinal studies following users from adolescence into adulthood.
A brain that is still forming new connections until the mid-twenties is not the same as a fully developed adult brain, and the research does not support applying adult risk profiles to adolescent use.
The declining perception of harm among young people, driven largely by normalisation through legalization and social media, is one of the more quietly serious public health developments of the past decade.
And it is one that the “cannabis is safer than alcohol” framing, whatever its merits for adults, may inadvertently be making worse.
The Honest Answer to an Honest Question
Is cannabis safer than alcohol?
For most adults using either substance moderately, the evidence suggests the answer is broadly yes, on most but not all measures of harm.
Alcohol carries a more clearly established cancer risk, causes more organ damage, kills more people directly, and contributes more to violence and accidents at high doses.
But cannabis is not risk-free, is potentially more addictive for regular users than most people realise, carries serious cardiovascular and respiratory risks at high frequency, poses real developmental dangers for users under 25, and is increasingly being used in forms and at potencies that bear little resemblance to the substance most of the existing research was conducted on.
The public perception that cannabis is safe has raced ahead of the evidence.
Not because cannabis is secretly as dangerous as alcohol, but because the cultural conversation has replaced one oversimplification with another rather than arriving at the more complicated truth.
That truth is worth knowing.
Both for the people who are making choices about their own health, and for the policymakers and public health professionals who will be managing the consequences of those choices for years to come.
Sources: YouGov, April 2024 | MRI-Simmons National Cannabis Study 2025, Marijuana Moment | Gallup Poll, August 2024 | NPR, May 2024 | CNN, May 2024 | American Heart Association, February 2024 | Healthline, June 2025 | US Centers for Disease Control and Prevention, Cannabis Health Effects | UCHealth Today | Kaiser Family Foundation
