Alcohol BAC Calculator
Estimate blood alcohol content (BAC) using the Widmark formula. Educational use only — NEVER use to decide if safe to drive. Uses WHO 2023 legal limits for Singapore, Malaysia, Thailand and six more countries.
Alcohol BAC Calculator
How to use the BAC Calculator
Read the warning first
This is an educational estimate. Widmark's model describes an average body, and a real blood alcohol level can differ from it a good deal: people eliminate alcohol at different rates, food changes how fast it is absorbed, and body composition changes how widely it spreads. Never use this tool to decide whether you can drive. If you have been drinking, arrange other transport.
Choose the drinks
Choose a drink and how many you had. Each preset is a volume and a strength, and the tool works out the grams of alcohol from those: millilitres × ABV × 0.7893, the density of ethanol. A 330 mL can of 5% beer is 13.0 g. For anything else, pick "Custom" and enter the grams yourself.
Enter the time, weight and sex
Enter the hours since your first drink, your weight and your sex. The model assumes all the alcohol was absorbed at once and has been eliminated since, so the clock starts at the first drink. The food question does not change the number: Widmark's model has no term for food. The tool tells you which way food would move a real result.
Choose where you are
Pick the country you are in. The legal limit is set by law, not by your body, and it runs from zero in Vietnam and Brunei to 0.08% in Singapore and the United States. The "hours until" figures assume the average elimination rate of 0.015% an hour and are rounded up to the minute. Some people are slower, so the result also shows the wait at 0.010% an hour, the slow end of the normal range. Treat either figure as a minimum, never as permission.
BAC — what it is, what it isn't, and why estimates miss
Blood alcohol concentration (BAC) is grams of alcohol per 100 mL of blood. The same number goes by several names: 0.08% is 0.08 g/dL, 80 mg per 100 mL and 0.8 g per litre. Erik Widmark worked out in the 1930s how to estimate it from a drinking history: divide the alcohol consumed by body weight times a distribution factor r, then subtract what the body has eliminated since. This tool uses Widmark's averages, r = 0.68 for men and 0.55 for women, and subtracts 0.015% for every hour. Real bodies differ: a 2015 study that measured r in 18 students found 0.71 to 0.81 in the men and 0.59 to 0.68 in the women (Sadler and Lennox, 2015).
Why it is an estimate and nothing more
The elimination rate is the weakest link. A 2010 review of the published measurements by Alan Jones found that 15 mg per 100 mL an hour, the 0.015% used here, is a good average for moderate drinkers. The same review found individual rates between about 10 and 35 mg per 100 mL an hour, the top of that range in alcoholics during detoxification. Two people who drink the same amount can therefore be hours apart in when they reach zero.
Absorption is the second gap. Widmark's model assumes the whole dose is in the blood at once. In real life food slows absorption, which lowers the peak and delays it. This tool used to multiply every empty-stomach result by 1.15 to stand in for that. The factor had no source, so it was removed on 30 September 2026; the food question now only tells you which way a real result would move.
"In moderate drinkers 15 mg/100mL/h remains a good average value for the population." Jones, Forensic Science International 2010
Forensic toxicologists use Widmark's method too, and they treat it with care. A 2007 review (Posey and Mozayani, 2007) notes that many practitioners are reluctant to estimate BAC this way and that some jurisdictions expressly prohibit it, and that published refinements to the formula are often ignored. It compares five models so that the spread between them is visible.
Legal limits in the region
Every limit in the country list comes from the World Health Organization's Global Information System on Alcohol and Health, using its 2023 data for general drivers. Singapore and the United States are 0.08%. Malaysia, Thailand and the Philippines are 0.05%, and Thailand sets 0.02% for young and novice drivers. Japan and South Korea are 0.03%. Vietnam and Brunei are zero. Malaysia shows why the date matters: the same WHO dataset listed it at 0.08% in 2019 and 0.05% in 2023. Laws change, so check the current rule before relying on any of these.
Singapore's limit is written into the Road Traffic Act 1961: section 72 sets the prescribed limit at 35 micrograms of alcohol in 100 mL of breath or 80 milligrams in 100 mL of blood, and section 67 makes it an offence to drive with more than that. The same section also makes it an offence to drive while so affected that you cannot keep proper control of the vehicle, whatever the reading. Being under the number is not a defence to that.
The alcohol flush and ALDH2
ALDH2 is the enzyme that clears acetaldehyde, the toxic product of the first step in breaking down alcohol. Many people in East Asia carry a variant that makes it work poorly, and when they drink, acetaldehyde builds up: the face flushes, the heart races and nausea follows. Writing in Brooks et al., PLoS Medicine 2009, Brooks and colleagues put the share at about 36% of East Asians (Japanese, Chinese and Koreans), and estimated at least 540 million ALDH2-deficient people worldwide, about 8% of the world's population. They also found that people with this response who drink regularly face a much higher risk of oesophageal cancer, and argued that the flush itself should be treated as a warning sign. The flush does not change the BAC figure. It changes what that figure means for your health.
Widmark's estimate, WHO's limits and the alcohol flush
Widmark's model estimates BAC from grams of alcohol, body weight, a distribution factor and time. It describes an average, not you.
Jones's 2010 review puts average elimination at 15 mg per 100 mL an hour in moderate drinkers, the 0.015% an hour this tool uses.
Individual elimination rates run from about 10 to 35 mg per 100 mL an hour, the high end in alcoholics during detoxification.
A US standard drink is about 14 g of pure alcohol (NIAAA); a UK unit is 8 g (NHS). The same can of beer is a different number of "drinks" depending on whose definition you use.
Grams come from volume, strength and ethanol's density of 0.7893 g/mL. A 330 mL can of 5% beer is 13.0 g, a little under one US standard drink.
WHO's 2023 data gives Singapore a general limit of 0.08%. Malaysia, Thailand and the Philippines are 0.05%, Japan and South Korea 0.03%.
Vietnam and Brunei set their limit at zero. Any measurable alcohol is over it.
Malaysia's limit was 0.08% in WHO's 2019 data and is 0.05% in the 2023 data. A limit on an old web page can be out of date.
About 36% of East Asians have the alcohol flushing response of ALDH2 deficiency, according to Brooks and colleagues in 2009.
Flushers who drink regularly have a much higher risk of oesophageal cancer, which is why Brooks and colleagues treat the flush as a warning.
Frequently Asked Questions
-
No. Widmark's model gives an average. Elimination alone varies more than threefold between people, from about 10 to 35 mg per 100 mL an hour in Jones's 2010 review, and the model ignores food. A forensic review in 2007 noted that some jurisdictions expressly prohibit estimating BAC this way. If you have been drinking, do not drive: take Grab, a taxi or public transport.
-
Food slows the passage of alcohol from the stomach to the small intestine, where most of it is absorbed. The peak comes later and is lower than on an empty stomach. Widmark's model cannot show this because it assumes the whole dose is absorbed at once, so its peak is at the high end of what you would really reach. This tool used to add 15% to every empty-stomach result; that factor had no source and was removed on 30 September 2026. The food question now only tells you which way a real result would move.
-
Alcohol spreads through the water in the body, and for the same weight women on average carry less body water than men. The same dose then reaches a higher concentration. Widmark's method captures this with the distribution factor r: this tool uses 0.68 for men and 0.55 for women. With the same drinks and weight, that gives a woman a peak about 24% higher. Jones's 2010 review found women's elimination rates slightly higher than men's on average, so the difference narrows a little over the hours.
-
ALDH2 is the enzyme that breaks down acetaldehyde, the toxic product of the first step in processing alcohol. A common variant in East Asia makes it work poorly, so acetaldehyde builds up and causes flushing, a fast heartbeat and nausea. Brooks and colleagues (PLoS Medicine, 2009) estimated that about 36% of East Asians have this response, and found that those who drink regularly face a much higher risk of oesophageal cancer. The BAC figure is the same either way; the health risk is not.
-
No. Once alcohol is in the blood, it leaves at the rate your liver clears it, and nothing on this list speeds that up. NIAAA warns that cold showers, hot coffee and walking do not reverse the effects of an alcohol overdose and can make things worse. Coffee can make you feel more alert, but your blood alcohol is unchanged. Only time brings the number down.
-
The only level with no alcohol in it is 0.00%. Legal limits are where each country draws the line, and they differ: 0.08% in Singapore and the United States, 0.05% in Malaysia and Thailand, 0.03% in Japan and zero in Vietnam and Brunei in WHO's 2023 data. A result under your local limit does not mean you are unaffected. It means this estimate, for an average body, falls under a legal threshold.
-
Per gram of alcohol, yes: ethanol is ethanol. What differs is how much a serving holds. A 330 mL can of 5% beer is 13.0 g, a 250 mL glass of 12% wine is 23.7 g and a 44 mL shot of 40% spirit is 13.9 g. The US NIAAA uses about 14 g as its standard drink and the UK's NHS counts units of 8 g, so that wine is 1.7 US drinks or 3.0 UK units. Strong craft beers, large wine pours and cocktails with more than one shot are where people most often underestimate what they drank.
-
This tool does not predict hangovers, and none of its sources measures them. What it can tell you is how much alcohol you drank and roughly when your blood level reaches zero. The way to have less of a hangover is to drink less.
-
No. The calculation runs in your browser, and what you type is never sent anywhere. The page still makes requests of its own (its scripts, an anonymous performance measurement, and ads or analytics only if you allow them), but none of them carries your drinks or your weight. Use it for awareness, never for a decision about driving.
-
The Caffeine Half-Life Calculator estimates how long coffee stays in your system, and the Sleep Calculator helps plan a bedtime. Neither makes you sober any faster. For getting home, use a ride-hailing app, a taxi or public transport.
Method & sources
What this tool implements
- Presets were typed grams (a 330 mL 5% beer as 14 g; it is 13.0 g) and an unsourced 'empty stomach ×1.15' factor inflated every result. Both are gone.
- Limits per WHO GISAH, indicator SA_0000001520, 2023 general-driver data, re-read 2026-10-16: Singapore and the United States 0.08, Malaysia 0.05 (0.08 in the same dataset for 2019), Thailand 0.05 and 0.02 for young drivers, Philippines 0.05, Japan and Korea 0.03, Vietnam and Brunei zero. Singapore's figure is also the statute: Road Traffic Act 1961 s 72 (35 µg/100 mL breath, 80 mg/100 mL blood) and s 67(1)(b) (exceeding it is the offence).
- The verdict is the law of the place chosen, never a clearance: no green or "sober" state exists, and a result under a limit says alcohol still affects you. NIAAA's 0.08% binge-drinking line triggers its overdose warning signs.
- BAC is shown in % (= g/dL), mg per 100 mL and g/L; grams as US standard drinks (about 14 g, NIAAA) and UK units (8 g, NHS).
- Corrected 2026-10-16: a green "✅ Sober" verdict below 0.01%, shown even under a zero limit (Vietnam, Brunei); an impairment ladder (0.03/0.05/0.08/0.15/0.30) with no source; "1h 60m" from rounding minutes to nearest (now rounded up); "over the limit in every country listed" hard-wired to 0.08 (now computed); "Open DevTools and confirm zero outbound requests" (the page makes its own requests; none carries what is typed); the Jones 2010 source title; Brooks 2009's 540 million is ALDH2-deficient people, "at least".
Sources
- Jones AW. Evidence-based survey of the elimination rates of ethanol from blood with applications in forensic casework. Forensic Sci Int 2010;200(1-3):1-20. https://doi.org/10.1016/j.forsciint.2010.02.021
- Posey D, Mozayani A. The estimation of blood alcohol concentration: Widmark revisited. Forensic Sci Med Pathol 2007;3(1):33-39. https://doi.org/10.1385/fsmp:3:1:33
- Coker AR, Weinstein DN, Vega TA, Miller CS, et al. Psychopharmacology 2020;237(10):3139-3148 — Widmark g = kg × r × BAC × 10 with r 0.68 males, 0.55 females, citing Posey & Mozayani. https://doi.org/10.1007/s00213-020-05599-5
- Sadler DW, Lennox S. Intra-individual and inter-individual variation in breath alcohol pharmacokinetics: variation over three visits. J Forensic Leg Med 2015;34:88-98. https://doi.org/10.1016/j.jflm.2015.05.020
- World Health Organization. Global Information System on Alcohol and Health: legal blood alcohol concentration (BAC) limits (GHO indicator SA_0000001520). https://ghoapi.azureedge.net/api/SA_0000001520
- Singapore Road Traffic Act 1961, sections 67 and 72 (prescribed limit). https://sso.agc.gov.sg/Act/RTA1961
- National Institute on Alcohol Abuse and Alcoholism. What is a standard drink? https://www.niaaa.nih.gov/alcohols-effects-health/what-standard-d…
- National Institute on Alcohol Abuse and Alcoholism. Understanding the dangers of alcohol overdose. https://www.niaaa.nih.gov/publications/brochures-and-fact-sheets/…
What can make this go out of date
- None at runtime; nothing the reader types is sent anywhere.
Abridged — the full review record for this tool runs longer than the list above.
Pick up where you left off
Stored only in this browser — never sent to our servers.