Roughly one standard drink an hour — but that is an average, the real range across people is wide, and nothing you can do speeds it up. Your liver clears alcohol at a fixed rate that does not care about coffee, cold showers, exercise or water.
This page gives you the actual numbers, the arithmetic behind them, how long each kind of test can detect alcohol, and the one safety point that matters more than any of it.
The number, and the range around it
The figure usually quoted is 0.015% blood alcohol per hour — the amount by which your blood alcohol concentration (BAC) falls every hour once you stop drinking.
The honest version is a range. The forensic toxicology consensus standard puts elimination at 0.010 to 0.025% per hour, a span that covers the large majority of people regardless of age, sex, or drinking experience. A slow eliminator takes two and a half times as long as a fast one to clear the same dose.
Two things about that rate are worth understanding, because they explain almost everything else on this page.
It is constant. Alcohol is cleared by liver enzymes that run at capacity, so the rate does not rise when you drink more — the NIAAA describes metabolism proceeding "at a steady rate, regardless of how much alcohol a person drinks." Ten drinks do not clear faster than three. They clear at the same speed for longer.
It is yours. The rate is set by your enzymes, and nothing you swallow, sweat out or splash on your face changes it.
Blood alcohol concentration (BAC) — grams of alcohol per 100ml of blood, expressed as a percentage. 0.08% is the legal driving limit in most of the US.
What counts as one drink
The NIAAA standard drink is 14 grams of pure alcohol:
- 12 oz of 5% beer
- 5 oz of 12% wine
- 1.5 oz of 40% spirits
Real pours run over. A craft IPA at 7–9% is a drink and a half. A generous restaurant glass of 14% wine is nearly two. A double is two. Most people undercount by a wide margin, and the arithmetic below is only as good as the count going in.
The arithmetic: how long until zero
Peak BAC depends on how much alcohol went in and how much water it had to dissolve into. The standard estimate uses body weight and a distribution factor — roughly 0.68 for men and 0.55 for women, because women carry proportionally less body water at the same weight.
Peak BAC ≈ grams of alcohol ÷ (weight in kg × factor × 10)
Hours to zero ≈ peak BAC ÷ 0.015
| drinks | 70kg man | 60kg woman |
|---|---|---|
| 1 | ~0.03% → about 2 hours | ~0.04% → about 3 hours |
| 3 | ~0.09% → about 6 hours | ~0.13% → about 8.5 hours |
| 5 | ~0.15% → about 10 hours | ~0.21% → about 14 hours |
| 8 | ~0.24% → about 16 hours | ~0.34% → about 23 hours |
Read the caveats before you read the table twice. This assumes the alcohol was absorbed instantly, on an empty stomach, in a person eliminating at the exact midpoint of the range. Real elimination varies 0.010–0.025, so a real person's time-to-zero could sit 40% either side of these figures. It is an estimate of a shape, not a measurement of you.
It is not a way to decide whether you can drive. A breathalyser is.
What that table is actually telling you
Look at the bottom two rows.
Eight drinks for a 70kg man — a heavy but not exceptional night — is roughly sixteen hours to zero. If the last one went down at 1am, zero arrives around 5pm. At 8am he is at about 0.13%, well over the limit, with a night's sleep behind him and every subjective sense that he is fine.
That is the morning-after driving problem, and it has been measured. In a study of festival attendees, 15% were estimated to still be over the legal limit the next morning, 4% tested over it, and among those who knew they were probably over, 45% still felt safe to drive. That was an Australian study against a 0.05 limit, so the exact figures do not transfer to the US — the dynamic does.
Feeling sober is not the same as being sober. The second one is a number.
How long each test can detect it
Different tests look for different things over different windows, and they are used for different purposes.
| test | what it detects | window | typical use |
|---|---|---|---|
| Breath | Alcohol still in blood | Hours — tracks BAC in real time | Roadside, DUI |
| Blood | Alcohol directly | ~1–12 hours | DUI evidence, medical |
| Urine (ethanol) | Alcohol directly | ~12–24 hours, longer after heavy use | General screening |
| Urine (EtG) | A metabolite left after alcohol is gone | Several days | Abstinence monitoring: courts, treatment, licensing |
| Saliva | Alcohol in real time | Parallels breath | Point-of-care screening |
| Hair (EtG) | Metabolite laid down as hair grows | ~90 days per segment | Long-term abstinence monitoring |
The one to understand is EtG. It is not alcohol; it is a residue your body makes from alcohol, and it lingers after the alcohol itself is gone — which is why it is the test used to check whether someone has been drinking at all, not whether they are drunk now.
It is also the test with a well-documented false-positive problem. Hand sanitiser, mouthwash, some foods and cleaning products contain enough alcohol to register, and SAMHSA's own advisory states that legal or disciplinary action based solely on a positive EtG result is "inappropriate and scientifically unsupportable." Peer-reviewed work has shown normal hand sanitiser use produces measurable EtG. If a result like this ever matters to you, insist on mass-spectrometry confirmation rather than the screening assay alone.
What does not speed it up
All of the following are widely believed and none of them change the elimination rate, because the rate is set by enzyme capacity in your liver and nothing on this list touches that:
- Coffee. Makes you a more alert drunk. The NIAAA lists it as a myth.
- Cold showers. Wake you up. Change nothing in your blood.
- Exercise, or "sweating it out." You lose trivial amounts of alcohol through sweat.
- Water. Worth drinking for other reasons. Does not accelerate clearance.
- Eating afterwards. Once the alcohol is absorbed, food does nothing to the rate.
The one thing that genuinely moves the numbers is food before and during drinking — it slows absorption, so the same drinks produce a lower peak. It does not change how fast the alcohol leaves once it is in. What to eat, and why it works.
What actually changes the numbers
- Body weight and composition. More body water dilutes the same dose to a lower BAC. This is most of why the men's and women's columns differ.
- Sex. Proportionally less body water in women at the same weight, per the NIAAA — a dose effect, not a resilience one.
- Genetics. The enzymes that clear alcohol and acetaldehyde come in variants. The best known is the ALDH2 variant behind the alcohol flush reaction, which slows acetaldehyde clearance and — the NIAAA is direct about this — raises certain cancer risks in people who drink through it.
- Regular heavy drinking. Induces a second clearance pathway; daily drinking measurably raises its activity within a week, and it falls back within about a week of stopping. This is tolerance at the enzyme level. It is not a good thing.
- Liver health, age and medications plausibly shift the rate too, through enzyme capacity and blood flow, though the quantitative data are thinner than for the factors above.
When it is not a hangover
Alcohol leaves on a fixed schedule. Two things that do not are worth recognising, because both get mistaken for a bad morning.
Alcohol poisoning is an overdose in progress — it happens while the alcohol is still in the table above, not after. If someone cannot be woken, is breathing fewer than eight times a minute or with long gaps, or has gone clammy or bluish, the NIAAA is blunt: do not assume they will sleep it off, do not leave them, call emergency services.
Withdrawal is what a body adapted to regular heavy drinking does when the drinking stops, and it can begin within the same window this page is about — shaking, sweating and anxiety that worsen rather than ease. That is medical, not a hangover. The full list of what a hangover is not.
Frequently asked questions
How long does alcohol stay in your system? Blood alcohol falls by roughly 0.015% an hour, within a documented range of 0.010–0.025. One standard drink clears in about one to two hours; a heavy night can take most of the following day. Metabolite tests can detect drinking for days afterwards.
How long after drinking can I drive? Only a breathalyser can tell you. The estimate table above shows why a heavy night commonly leaves people over the limit well into the next morning while feeling fine.
Does coffee sober you up? No. It makes you more alert; it does not change your blood alcohol at all.
How long does alcohol show in a urine test? Alcohol itself, roughly 12–24 hours. The EtG metabolite, several days — with a false-positive problem serious enough that SAMHSA warns against acting on a screening result alone.
Does food slow down alcohol? Food before and during drinking slows absorption and lowers the peak. Food afterwards does nothing to the rate at which it leaves.
Does drinking water flush alcohol out faster? No. Elimination is enzyme-limited. Water is worth drinking for the fluid you are losing, not for the alcohol you are clearing.
Where Snake Oil fits
Nothing on this page is something a drink mix can change, ours included. The clearance rate is your liver's, and we would not want to be the company that told you otherwise.
What the hours in that table do involve is fluid and mineral loss — alcohol turns off the kidney's hold-water signal, and sodium, potassium and magnesium leave with the water. That part is replaceable, and it is what we built for. 370mg of sodium and the 5g of carbohydrate that lets your gut absorb it, plus potassium, magnesium, zinc, the full B lineup, vitamins C and E and a glutamine-led amino acid blend — one stick, 16 ounces.†
Take it while you wait. The waiting is not optional.
†These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
Related reading: How long does a hangover last? · Does alcohol dehydrate you?