For the first time in the eight decades Gallup has been asking Americans what they put in their lungs, the answer has quietly flipped. More U.S. adults now say they smoke marijuana than say they smoke cigarettes — and the gap between the two is the widest it has ever been.
The numbers come from Gallup’s annual Consumption Habits poll, published on August 24, 2026 and based on interviews conducted between July 1 and July 19. You can read the full release here: U.S. Marijuana Smoking at Record High, Cigarettes at Low. Seventeen percent of American adults told Gallup they smoke marijuana, tying the all-time high the pollster first recorded in 2023. Eleven percent said they smoke cigarettes — matching the lowest figure in a trend line that stretches back to 1944, when the answer was 41%.
Put those two lines on the same chart and you are looking at one of the strangest reversals in modern American life: a legal, heavily taxed, government-warned-about product collapsing, and a formerly criminal one climbing past it.
Key Takeaways
Seventeen percent of U.S. adults now report smoking marijuana, more than double the 7% Gallup measured in 2013 and level with the 2023 record. Cigarette smoking sits at 11%, its joint-lowest reading ever and a quarter of the 1954 peak of 45%. Separately, 15% of Americans say they eat cannabis edibles — the highest of the three readings Gallup has taken on that question. Use skews young, with a quarter of adults aged 30 to 49 smoking cannabis compared with 5% of those 65 and over. Yet public opinion has not followed consumption: Americans split almost exactly down the middle on whether marijuana helps or harms the people who use it, 47% to 46%. And a caution worth carrying through the whole article: these are self-reported answers from a sample of 1,200 people with a margin of error of about four percentage points, which means the single-year jumps are far less certain than the decade-long trend.
The Number That Flipped
The headline comparison is genuinely striking. In 2013, when Gallup first asked its current marijuana-smoking question, 7% of adults said yes and roughly 19% said they smoked cigarettes. Thirteen years later those lines have not just converged — they have crossed and kept going. Marijuana smoking has more than doubled. Cigarette smoking has nearly halved.
Cigarettes have been falling for a very long time. Gallup’s tobacco trend begins in 1944 at 41%, peaks at 45% in 1954, drops below 30% for the first time in 1989, and has sat below 20% continuously since 2015. What’s new is not the decline itself but where it has landed: for five consecutive years now, the figure has hovered between 11% and 12%, which starts to look less like a slide and more like a floor made of the people who are not going to quit.
Marijuana’s rise is younger and much bumpier. The trend goes 7% in 2013, 13% in 2016, 12% for several years, then 16% in 2022 and 17% in 2023 — before dropping back to 13% in 2024, recovering to 14% in 2025, and hitting 17% again now. That zigzag matters, and it is the first thing an honest reading of this poll has to deal with.
Why “Record High” Deserves an Asterisk
Here is the part most coverage will skip. Gallup interviewed 1,200 adults by telephone, and the margin of sampling error is plus or minus four percentage points. That means the difference between this year’s 17% and last year’s 14% is inside the range where random sampling noise alone could explain it. The same applies to the 2024 dip. The safe interpretation is not “marijuana use surged this summer” but “marijuana use has roughly doubled over thirteen years and is currently bouncing around the high teens.”
There is a second, subtler issue: everything here is self-reported. People are being asked, over the phone, by a stranger, whether they use a substance that is fully legal in some states, medically restricted in others, and still federally controlled. As legalization spreads and social stigma fades, some of the measured increase is almost certainly people becoming more willing to say yes — not people newly starting. Researchers call this social desirability bias, and it cuts in the opposite direction for cigarettes, which have gone from glamorous to slightly shameful over the same period.
None of this makes the trend fake. It makes the trend real and the precise annual numbers soft. That distinction shows up clearly in other datasets too: a long-running study of what actually changed in California after seven years of legal dispensaries found that overall use rose far less dramatically than either supporters or opponents predicted, while the way people consume changed enormously.
The Edible Boom Hiding Behind the Headline
The most interesting number in this poll is not the one in the title. Fifteen percent of U.S. adults now say they eat cannabis edibles — gummies, chocolates, drinks, baked goods. Gallup has asked this only three times: 14% in 2022, 12% in 2024, and 15% now.
Add that to the smoking figure and the picture changes. Cannabis is no longer primarily a smoked product in the American imagination; it is a category with two roughly equal delivery formats, one of which requires no lighter, produces no smell, and looks like candy.
The demographic split is revealing. Among smokers, men slightly outnumber women, 18% to 15%. Among edible users the pattern reverses: 17% of women versus 12% of men. Income tells a similar story. Marijuana smoking falls steadily as household income rises — 24% among those earning under $50,000, 13% among those above $100,000 — while edible use stays essentially flat across every income bracket, at 17%, 16% and 16%.
That flatness is the tell. Edibles are not tracking the socioeconomic pattern of smoking. They are behaving like a mainstream consumer product, which is roughly what the beverage industry has been betting on with its projections that cannabis drinks will become a nine-billion-dollar category by 2036.
Who Is Actually Using: Age, Wallet, and Church Attendance
The age gradient is steep and consistent. A quarter of adults aged 30 to 49 smoke marijuana (25%), as do 23% of those aged 18 to 29. Then it falls off a cliff: 10% among 50-to-64-year-olds and 5% among those 65 and older. Edible use follows the same shape — 20% and 19% in the two younger groups, 12% and 7% in the older ones — but slightly less sharply, which fits the idea that edibles are the more approachable entry point.
Education splits the sample too: 20% of adults without a college degree smoke cannabis, against 11% of graduates.
The strongest single predictor in the whole table, though, is religious attendance. Among Americans who seldom or never attend religious services, 22% smoke marijuana and 20% eat edibles. Among those who attend weekly, both figures drop to 4%. That is a fivefold gap — bigger than age, bigger than income, bigger than education. Whatever is driving American cannabis use, it is entangled with culture and community far more than with any single demographic variable.
Alcohol is the other strong companion signal: 21% of drinkers smoke marijuana, versus 12% of non-drinkers. The two substances travel together rather than substituting for one another, at least at the level of individual behavior.
Nicotine Didn’t Die. It Changed Shape.
If you only read the cigarette line, you would conclude that Americans are giving up nicotine. They are not. They are switching containers.
Nine percent of adults now vape, a figure that has stayed in a narrow 6%-to-9% band since Gallup started measuring it in 2019. And for the first time, Gallup asked about nicotine pouches — the small white sachets tucked under the lip, sold under brand names like Zyn. Four percent of American adults said they had used one in the past week.
The age patterns diverge sharply from cigarettes. Vaping peaks among 18-to-29-year-olds at 17% and falls to 1% among those over 65. Cigarette smoking, by contrast, peaks in middle age at 16% among 30-to-49-year-olds and is only 8% among the youngest adults. Nicotine pouches are overwhelmingly male: 7% of men, 1% of women.
Pouches also break the usual pattern in one telling way. Cigarettes and vapes are both more common among lower-income Americans; pouch use runs the other direction, rising with income to 6% among households above $100,000. It is the only nicotine product in the poll that skews affluent.
Americans Still Can’t Agree Whether Any of This Is Good
Consumption has doubled. Opinion has barely budged.
Asked whether marijuana has a positive or negative effect on most people who use it, 47% of Americans said positive — 9% very, 38% somewhat — and 46% said negative, split 16% very and 30% somewhat. That is a statistical dead heat. It is also a slight rebound from 2024, when opinion leaned mildly negative at 51%, and a slight retreat from 2022, when it leaned mildly positive at 53%.
What that means in practice is that a country where roughly one adult in six uses cannabis is a country where roughly half of adults think it is bad for the people who use it. Those two facts sit side by side without resolving, which helps explain why American cannabis policy remains such a patchwork. It also fits what surveys of users themselves consistently find about the reasons Americans actually give for using cannabis — motivations that blend medical, social and recreational purposes in ways that public opinion polling struggles to capture with a single up-or-down question.
The Bigger Picture: Weed Up, Booze Down
One more finding gives the whole poll its context. In the same survey cycle, Gallup found that 54% of Americans drink alcohol — a record low, holding there for a second consecutive year.
So within a single dataset: alcohol at its lowest ever, cigarettes at their lowest ever, marijuana at its highest ever, edibles at their highest ever, and nicotine pouches appearing in the data for the first time. This is not a story about Americans consuming more or less. It is a story about substitution, generational turnover, and a shift in what counts as a normal, unremarkable thing to do on a weekday evening.
That shift is not confined to the United States. Global monitoring bodies now estimate that cannabis use worldwide has reached roughly 256 million people, with legal and illegal markets interacting in ways that no single national poll can capture. The American numbers are one visible slice of a much larger change — the slice where the data happens to go back eighty-two years.
Frequently Asked Questions
Probably not in the way the word “spike” suggests. With a sample of 1,200 people and a margin of error of about four percentage points, the move from 14% last year to 17% this year falls within the range that random sampling variation can produce on its own. The reliable finding is the long arc: 7% in 2013 to the high teens now.
They are different products facing opposite social and legal currents. Cigarettes have absorbed sixty years of health campaigns, taxation, indoor bans and cultural stigma. Cannabis has moved the other way — toward state-level legalization, medical availability and reduced stigma. Some of the measured increase may also reflect people becoming more willing to admit use rather than more people using.
They avoid the harms associated with inhaling combusted plant material, which is a genuine difference. But they carry their own risks, particularly around dosing: effects arrive slowly and can be much stronger and longer-lasting than expected, which is a common route to accidental overconsumption. Neither format is risk-free, and this poll measured behavior, not health outcomes.
No. Gallup surveyed only U.S. adults, and cannabis prevalence varies enormously by country depending on legal status, enforcement, cultural attitudes and how surveys are conducted. Treat these numbers as a portrait of the United States in July 2026, not a global benchmark.
Disclaimer
This article is provided for informational and educational purposes only. It reports the findings of a public opinion survey and does not constitute medical, health or legal advice, nor an endorsement or encouragement of the use of any substance. Cannabis laws differ substantially between countries and, in some cases, between regions within the same country. Readers should consult a qualified healthcare professional regarding any health question and should inform themselves of the legislation in force in their own jurisdiction. The Cannex accepts no liability for decisions taken on the basis of this content.
