BAC Calculator

No. This is an educational estimate and not a breath, blood, legal, or medical measurement. The only safe driving BAC is 0.000%, and impairment can occur before a legal limit is reached.

No. A breathalyzer estimates alcohol from a breath sample. This calculator estimates blood alcohol concentration from drink timing, drink size, ABV, body weight, food, and model assumptions. It can help visualize a possible BAC timeline, but it is not a breathalyzer, breath alcohol test, blood test, or legal measurement.

A BAC calculator estimates blood alcohol concentration from entered assumptions, while a breathalyzer estimates alcohol from a breath sample at the time of testing. A breathalyzer is a measurement device; this calculator is a forward simulation based on the drinks, timing, food, and body information you enter.

The two can differ because breath alcohol readings, blood alcohol concentration, individual physiology, drinking pattern, recent food, and measurement conditions are not identical.

This blood alcohol calculator uses a Modified Norberg pharmacokinetic model for ethanol. The model includes gastrointestinal absorption, a liver compartment, central and peripheral distribution compartments, and Michaelis-Menten elimination. Zekan et al. adapted a Norberg-style model for oral ethanol administration and reported the population parameter estimates used here.

A classic Widmark formula is a rough screening calculation: it typically treats alcohol as quickly distributed into a single body-water compartment and subtracts a linear elimination rate. The Modified Norberg approach is a substantially more physiologically detailed model. It represents gastrointestinal absorption, liver elimination, and distribution between compartments over time.

Because it models processes that the Widmark formula simplifies or ignores, the Norberg-style approach is generally better suited for estimating the shape of the BAC curve, peak BAC, and peak timing. It can be much more informative than a simple Widmark estimate, especially when drink timing, food, and delayed absorption matter. That still does not make it a guaranteed prediction of one person's true BAC, because physiology, inputs, and measurement conditions vary.

Yes. Food and gastric emptying can strongly affect alcohol absorption, peak BAC, and the time it takes to reach peak blood alcohol content. Food entries in this calculator decay over time as gastric load and modify only the existing Norberg absorption terms kA and f2; they do not create a separate stomach-intestine PBPK model.

The cited Modified Norberg study was fitted in healthy young male volunteers. The male/default option uses the published distribution-volume constants. The female option lowers the distribution volume to approximate lower average total body water, but it is an extrapolation, not a separately fitted female BAC model.

In the United States, one standard drink contains about 14 grams, or 0.6 fl oz, of pure alcohol. That is approximately 12 fl oz of 5% beer, 5 fl oz of 12% wine, or 1.5 fl oz of 40% spirits. The NIAAA explains the U.S. standard drink definition in more detail here.

BAC is shown here as a percent value. To convert BAC percent to promille, multiply by 10. For example, 0.05% BAC is about 0.5‰, and 0.08% BAC is about 0.8‰.

Promille is commonly used in continental Europe and in German terms like Promillerechner. The conversion only changes the displayed unit; it does not change the uncertainty of the BAC estimate.

It can show how assumptions change a BAC curve, but it cannot predict an individual's true BAC with forensic precision. The model parameters, drink entry, physiology, food, health status, and measurement method all introduce uncertainty.

The interval reflects uncertainty in selected published model parameter estimates under the chosen drink and food scenario. It is not a prediction interval for your true BAC. It excludes individual variation, medication effects, liver function, tolerance, drink-entry error, breath or blood testing error, and legal evidence standards.