The formula behind the result
The Swedish chemist Erik Widmark published this method in 1932, and it is still the basis of most blood alcohol estimates. It treats the body as a single container: the alcohol consumed is spread through the part of the body that holds water, and the liver removes it at a roughly constant rate.
BAC (%) = A ÷ (r × W × 10) − β × t
- A is the alcohol consumed in grams: the number of standard drinks multiplied by the grams of alcohol in each.
- W is body weight in kilograms.
- r is the Widmark distribution factor, the fraction of body weight through which alcohol spreads. Widmark’s average values were 0.68 for men and 0.55 for women.
- 10 converts grams per kilogram into grams per 100 millilitres, the unit written as a percentage.
- β (beta) is the elimination rate in percentage points per hour. The calculator starts at 0.015, a commonly used average, and lets you change it.
- t is the time in hours since the first drink.
If the subtraction gives a negative number, the estimate is shown as zero.
What a standard drink is
The formula needs grams of pure alcohol, and “a drink” is not a fixed amount. In Australia a standard drink is defined as 10 g of alcohol, and packaged drinks state the number of standard drinks on the label. Other countries define it differently: a United Kingdom unit is 8 g and a United States standard drink is 14 g. That is why the grams per drink can be edited. A restaurant glass of wine or a pint of full-strength beer usually holds more than one Australian standard drink.
Worked example
A man who weighs 80 kg has four Australian standard drinks, and two hours have passed since the first.
- Alcohol consumed. 4 × 10 = 40 g.
- Before elimination. 40 ÷ (0.68 × 80 × 10) = 40 ÷ 544 = 0.074%.
- Eliminated. 0.015 × 2 = 0.030 percentage points.
- Estimate. 0.074 − 0.030 = 0.044% (0.0435 before rounding).
Change only the elimination rate and the answer moves a long way: at 0.010 per hour the estimate is 0.054%, and at 0.020 it is 0.034%. A woman who weighs 60 kg, with the same drinks and time, gets 40 ÷ (0.55 × 60 × 10) = 40 ÷ 330 = 0.121%, less 0.030, which is 0.091%. The same four drinks give estimates from 0.034% to 0.091% across these ordinary cases, and none of them is a measurement.
Why the estimate can be far from a real reading
Every input is uncertain. The distribution factor is an average for a whole sex; it varies with body fat, muscle, age and hydration, and two people of the same weight can differ substantially. Elimination rates differ between people and between occasions. People commonly miscount drinks, and home pours are often larger than a standard drink.
The formula also assumes the alcohol is absorbed instantly. In reality absorption takes time and depends heavily on food, so concentration can still be rising well after the last drink. An estimate made shortly after drinking can be too high or too low, and a person can feel fine while their concentration is climbing.
For these reasons the number is useful only for understanding how the factors interact. It is not evidence of anything, it cannot stand in for a breath test, and it says nothing about impairment, which begins at low concentrations and differs from person to person. Legal limits vary by place, licence type and vehicle; check with your local road authority, and do not use this page to judge where you stand against them. If you have been drinking, do not drive.
If you are worried about your drinking or someone else’s, a doctor can help, and so can an alcohol and drug support line in your state or territory.