For twenty years, anyone in the United States who asked "how much is too much?" got a number back. Two drinks a day for men, one for women. It was printed in the Dietary Guidelines for Americans, quoted in doctors' offices, and used as the invisible yardstick millions of people measured themselves against. If you stayed under it, you were fine. That was the deal.
On January 7, 2026, the number disappeared. The 2025-2030 Dietary Guidelines for Americans dropped the daily limits entirely and replaced them with a single sentence of advice: consume less alcohol for better health. No thresholds, no per-day cap, no distinction between men and women. Several medical bodies, including the American Association for the Study of Liver Diseases, publicly objected to the removal of evidence-based numbers.
Which leaves the original question wide open, and for a lot of people, uncomfortably so. Because "drink less" is directionally right and practically useless. Less than what? The yardstick people had been quietly grading themselves against is gone, and the honest reaction to that is not relief. It is the realization that most of us never knew what the number meant in the first place.
Here is what a standard drink actually is, why the glass in your hand is almost certainly more than one, what limits still exist elsewhere in the world, and how to work out where you personally land.
First, the Unit Nobody Measures In
Every official drinking guideline in the world is written in a unit almost nobody uses at home. In the United States, one standard drink contains 14 grams of pure alcohol, which is about 0.6 fluid ounces of ethanol. That works out to:
- 12 oz of beer at 5% ABV
- 5 oz of wine at 12% ABV
- 1.5 oz of spirits at 40% ABV
Notice how specific those are. Not "a beer." Not "a glass of wine." A precise volume at a precise strength. And this is where the entire system quietly breaks, because real drinks have drifted a long way from those numbers.
Run the arithmetic on what people actually pour:
- A 16 oz pint of craft IPA at 6.5% is about 1.7 standard drinks
- A 16 oz pint of a stronger double IPA at 9% is about 2.4 standard drinks
- A generous 9 oz pour of modern red wine at 14% is about 2.1 standard drinks
- A 750 ml bottle of wine at 13.5% holds about 5.7 standard drinks
- A home spirits pour of 2.5 oz, which is what most people free-pour, is about 1.7 standard drinks
Now apply that to the most common self-description in the English language: "I have a couple of glasses of wine with dinner." Two real-world 9 oz pours of 14% wine is roughly 4.2 standard drinks. Every night, that is around 29 standard drinks a week, from someone who would sincerely tell their doctor they drink moderately. They are not lying. They are using a different unit than the one the guidelines are written in.
This is the single biggest reason people misjudge their own intake, and it is worth sitting with before reading any threshold below. A guideline in standard drinks is meaningless until you convert your actual glass. The math is simple: multiply the volume in ounces by the ABV percentage, then divide by 0.6. Anything above 1.0 means your one drink was never one drink.
The Thresholds That Still Exist
The Dietary Guidelines dropped their numbers, but the National Institute on Alcohol Abuse and Alcoholism did not. NIAAA still publishes the definitions used across American clinical practice and research, and they are worth knowing precisely.
Binge drinking is a pattern that brings blood alcohol concentration to 0.08% or above. For a typical adult that means about 5 or more drinks for men, or 4 or more for women, within roughly two hours. The two-hour window matters: the same five drinks spread across an evening is a different physiological event than five drinks in two hours.
Heavy drinking is defined for men as 5 or more drinks on any day, or 15 or more per week. For women it is 4 or more on any day, or 8 or more per week.
Those weekly numbers deserve a second look, because they are where most people find themselves without ever having a dramatic night. A woman having two real glasses of wine four evenings a week is over the threshold. Not from binges, not from blackouts. From Tuesdays.
Why Every Country Gives a Different Answer
If you go looking for the "real" limit internationally, you find striking disagreement. Here is the current guidance, with everything converted into US standard drinks so the comparison is fair:
| Authority | Guidance | In US standard drinks |
|---|---|---|
| US 2020-2025 (retired) | 2/day men, 1/day women | ~14/week men, ~7/week women |
| US 2025-2030 (current) | "Consume less alcohol for better health" | No number given |
| Canada (CCSA, 2023) | 2/week low risk; 3-6 moderate; 7+ increasingly high | ~2/week |
| UK (Chief Medical Officers) | 14 units/week, both sexes, spread over 3+ days | ~8/week |
| Australia (NHMRC, 2020) | 10/week, and no more than 4 on any day | ~7/week |
| WHO / IARC | Group 1 carcinogen; no safe level established | 0 |
The spread is enormous. Canada's low-risk threshold is about two drinks per week. The retired American guideline for men was about fourteen. That is a sevenfold difference between two wealthy countries reading the same body of evidence in the same decade.
The disagreement is not really about the science. It is about how much risk each body decided to treat as acceptable, and how honestly they chose to communicate it. Canada's 2023 guidance abandoned the idea of a safe threshold and instead published a risk continuum: zero is the only no-risk option, 1 to 2 drinks per week is low risk, 3 to 6 is moderate, and 7 or more climbs steeply. The UK settled on a level where lifetime risk of alcohol-related death stays under roughly 1%. Australia picked its own tolerance. None of them found a magic safe line, because there isn't one to find.
What collapsed in the background was the old J-curve, the comforting research finding that light drinkers outlived teetotalers. That result has largely been attributed to a study design flaw: the "never drink" comparison groups were contaminated with former drinkers who had quit because they were already sick. When later studies separated lifelong abstainers from sick quitters, the protective effect shrank or vanished. The heart-healthy glass of red wine was, to a significant degree, a statistical artifact.
What the Risk Curve Actually Looks Like
Once you drop the search for a safe line, the picture gets simpler and more honest. Alcohol risk is a dose-response curve that starts at zero and rises. There is no cliff edge where you suddenly become a problem drinker. There is a slope, and every drink moves you up it a little.
For cancer specifically, the slope starts earlier than most people expect. The US Surgeon General's January 2025 advisory laid this out directly: alcohol is causally linked to at least seven cancers, including mouth, throat, larynx, esophagus, liver, breast, and colorectal, and it contributes to roughly 20,000 cancer deaths a year in the United States. Breast cancer risk begins rising at levels as low as one drink per day. Alcohol has been classified by the WHO's International Agency for Research on Cancer as a Group 1 carcinogen, the same category as tobacco and asbestos, since 1988. We cover what the cancer evidence does and does not say in more detail, including the crucial distinction between relative and absolute risk that headlines routinely mangle.
Other systems have their own slopes at their own gradients. Blood pressure responds almost linearly, with roughly 1 mmHg of systolic pressure per daily drink. Sleep architecture degrades from the first drink of the evening. Liver risk stays quiet for years and then is not quiet at all.
So the honest answer to "how much is too much" is: too much for what? Too much to raise your cancer risk at all is zero. Too much to be in the lowest-risk band Canada recognizes is more than two a week. Too much to avoid a clinical heavy-drinking label is 8 or 15 a week depending on sex. Too much to avoid acute harm tonight is 4 or 5 in two hours. Those are four different questions with four different answers, and conflating them is what lets people reassure themselves with whichever number is most convenient.
How to Find Out Where You Actually Are
Rather than guessing, two exercises give you a real answer in about ten minutes.
Run the AUDIT-C on yourself. It is a three-question screen used in primary care worldwide, and it takes under a minute. How often do you have a drink containing alcohol? How many drinks do you have on a typical drinking day? How often do you have six or more drinks on one occasion? Each answer scores 0 to 4. A total of 4 or more for men, or 3 or more for women, is considered a positive screen worth discussing with a doctor. It is not a diagnosis. It is a flag.
Log one honest week. Not a typical week, not a good week. The next seven days, exactly as they happen, with every pour converted into standard drinks using the arithmetic above. Most people are genuinely surprised, and the surprise is almost never in the direction of "less than I thought." Estimating from memory reliably undercounts, because memory rounds down and forgets the top-ups.
If you want the number to keep meaning something after that first week, it has to be recorded as you go rather than reconstructed on Sunday night. Sober Tracker is built for exactly this kind of private, no-account tracking, whether you are counting alcohol-free days or just getting an accurate baseline before deciding anything.
One thing worth knowing before you start counting: the answer to "how long until it is out of my system" is not the same as "how long until I am unimpaired." We break that down in how long alcohol stays in your system, and it changes how you read your own weekly log.
When the Number Is Not the Point
There is a category of drinker for whom this entire article is beside the point, and it is important to name it clearly.
If you are drinking within guidelines but thinking about alcohol constantly, the number is not your problem. If you have repeatedly decided to cut back and repeatedly not managed it, the number is not your problem. If you drink to manage anxiety, or to sleep, or to get through evenings, the number is not your problem. Plenty of people who would pass a guideline check are stuck in grey area drinking, where nothing looks alarming from outside and something is clearly wrong from inside.
Guidelines measure exposure. They do not measure relationship. Someone drinking six a week while white-knuckling every evening is in more trouble than someone drinking eight a week and never thinking about it. This is also why "just cut down to the guideline" fails so often as a plan: for anyone with real dependence, moderation demands the exact executive function that alcohol has been eroding. We go into why moderation so often fails and what tends to work instead.
A safety note that matters more than any threshold here: if you have been drinking heavily and daily for a long time, do not stop abruptly on your own. Alcohol withdrawal can be medically dangerous, and the withdrawal timeline is worth understanding before you change anything. Talk to a doctor about tapering.
The Honest Conclusion
The United States removed its drinking limit because the evidence stopped supporting the idea of a comfortable safe zone, not because drinking got safer. The rest of the world responded to the same evidence by lowering their numbers rather than deleting them, which is why Canada now says two a week where America used to say fourteen.
What that leaves you with is not a permission slip and not a verdict. It is a slope, your position on it, and a choice about where you want to stand. Most people who go looking for the line are really asking whether they are okay. The threshold cannot answer that. What can is a week of honest counting, a converted pour, and a straight look at whether alcohol is something you have or something that has you.
If the counting turns out worse than you expected, that is not a failure. It is the first accurate piece of information you have had, and every version of this that ends well starts exactly there. A dry month is one of the cheapest experiments available, and the organ-by-organ recovery map shows what starts changing once the counting stops going up.
Getting an honest baseline before you decide anything? Sober Tracker is a private, no-account tracker with no social feed and no pressure. Count your alcohol-free days, or just find out what a normal week really looks like.
This article is educational and not a substitute for medical advice. Drinking guidelines are population-level tools and do not account for pregnancy, medications, liver disease, family history, or personal risk factors. If you are concerned about your drinking, talk with a healthcare provider. Sudden withdrawal from heavy long-term drinking can be dangerous and should be medically supervised.
